Mechanical Engineering | Physics |

Preservation Of The Architectural

Heritage | spatial planning and urban
development | Structural, Seismic
And Geotechnical Engineering | Urban
Planning, Design And Policy | Aerospace
Engineering | architectural composition
| Architecture, Built Environment And
Construction Engineering | architecture,
urban design, conservation of housing
and landscape | Architectural, Urban
And Interior Design | Bioengineering |
Design | Electrical Engineering | Energy
And Nuclear Science And Technology
| Environmental And Infrastructure
Engineering | Industrial Chemistry And
Chemical Engineering | Information
Technology | Interior Architecture and
Exhibition Design | Management Engineering
| Materials Engineering | Mathematical
Models And Methods In Engineering
PhD Yearbook | 2018

Doctoral program
in Industrial Chemistry and
Chemical Engineering
Coordinator: The Reference Committee
Prof. Alessio Frassoldati Margherita Albano (Solvay Specialty Polymers Italy SpA), Vincenzo Guida
(Procter & Gamble), Renato Paludetto (Dow Italia), Paolo Pollesel (ENI),
Gianmarco Polotti (Lamberti), Paolo Vacca (SAES-Getters).
Professors Committee
Alessio Frassoldati, Alessandra Beretta, Giulia Bozzano, Elisabetta Brenna,
Carlo Cavallotti, Attilio Citterio, Pio Forzatti, Francesco Gatti, Luca Lietti,
Davide Manca, Maurizio Masi, Andrea Mele, Pierangelo Metrangolo,
Massimo Morbidelli, Davide Moscatelli, Walter Navarrini, Isabella Nova,
Roberto Piazza, Carlo Punta, Giuseppe Resnati, Selena Sironi, Giancarlo
Terraneo, Enrico Tronconi.
Grants
Pirelli, Captive Systems, Humanitas, EU H2020-MSCA-ITN-2014.
The Doctoral Program in Industrial Chemistry and Chemical Engineering (CII)
is designed for students aiming to get a deep expertise on chemical processes
and on material processing either as the inherent research or in the process
design and development activities. The program is the ideal extension of the
Laurea Magistrale degrees in Chemical Engineering, Safety and Prevention
Engineering, Materials Science, Material Engineering, Industrial Chemistry and
Chemistry, but it is also open to graduated in other scientific disciplines.
Structure simulation of Amyloidal peptide nanohelices in
Nitrogenase enzyme for N2 solution as ordered nanomaterials
conversion to NH3 (Nature Commun. 2015, 6:7574, DOI:
10.1038/ncomms8574)
The general topic for the doctoral program in Industrial Chemistry
and Chemical Engineering is the application of the chemical and physical
knowledge to all the activities related to the synthesis, design, production
PhD Yearbook | 2018
and transformation of chemical substances and materials (like advanced
inorganic, polymers and biomaterials). These studies involve not only the
synthesis processes but also the related plants, here analyzed starting from
the laboratory tests and the pilot plant experiments up to the industrial size
ones, always including the careful evaluation of the related energetic, safety
and environmental issues. The final goal is to provide to the PhD student
the tools and the skills: (1) to design and manage industrial processes at any
size scales; (2) to develop new technical applications and (3) to create and to 315
characterize new products and services. The CII program covers three years
Industrial Chemistry and Chemical Engineering
for an overall amounts of 180 credits. The responsibility of the organization
and of the contents of the doctoral program is attributed to the Professors
Committee (PC). At the beginning of the Doctoral program, a tutor and
a thesis advisor is assigned to each student. The tutor has a supervisor
function during the whole doctoral program, whereas the thesis advisor is
responsible for the thesis work.
“Biocompatible polymer nanoparticles VLE and VLLE measurements for the
functionalized with a fluorescent dye system CH4-H2S-H2O.
for in vivo imaging studies”
The PC defines a specific curriculum for each student, on the basis of the
tutor suggestions. The curriculum has to be designed in order to both cover
deficiencies in the student scientific formation and to give a high-quality
technical and scientific preparation. Since 2001 (XVII cycle) the CII program
graduated more than 100 students, being ~80% now working in industry
and the remaining in the university or government research centers. These
data demonstrate the great link existing between the performed researches
and the industry. In fact, during the years, numerous the research topics
were directly supported by industrial companies, like ENEL, LPE Epitaxial
Technology, Bracco, Flamma, Mapei, Solvay Specialty Polymers, RSE, Isagro,
Tecnimont, Biochemtex, Pirelli.
Simulation of a chemical vapor Hazardous gas dispersion simulation
deposition reaction for silicon films from an industrial accident
deposition
 316
Chemo-enzymatic depolymerisation
of lignins
Chiara Allegretti - Supervisor: Prof. Paola D’Arrigo
The experimental work performed
in these three years of PhD,
locates within the broad field of
lignin valorization studies. Lignin
represents more than 30 % (by
weight) of the total biomass of the
earth biosphere and it is one of
the three major subcomponents
of lignocellulosic biomass.
Although burning lignin still
represents a valuable contribution
for saving fossil sources, lignin also
offers perspectives in terms of
higher value-added applications.
In fact, after the hydrolysis of
lignocellulose polysaccharides,
lignin remains as a solid residue,
representing one of the major
renewable sources for biofuels
and fine chemicals (mainly
aromatics) production. The main
limitation to practical applications
for lignin so far is presumably due
to the difficult and challenging
processing for depolymerisation
and valorisation. In this respect,
the challenge is to combine
chemical and enzymatic methods
in order to obtain selective
cleavages of the polymeric
skeleton. The work of my PhD is
inserted in the context of the Valor
Plus European Project “Valorisation
of biorefinery by-products
leading to closed loop systems
with improved economic and
environmental performance” (FP7
EC-KBBE-CALL-7 grant agreement
FP6-KBBE-2013-7-613802
(2013-2017)) that aims to set-up
a chemo-enzymatic process for
the controlled depolymerisation
and transformation of standard
lignin feedstocks to value product
streams.
The work accomplished in this
PhD project is divided into three
main sections: Part A - The Lig
enzymatic system, Part B - Lignin
fractionation, Part C - Lignin
depolymerisation.
At first, the Lig enzymatic system
(Part A) was studied, which is
able to cleave the β-O-4 aryl
ether linkage, which represents
about 50 % of all ethers in various
lignins. The cleavage of this type
of bond was identified to be
part of lignin catabolism in the
proteobacterium Sphingobium
sp. SYK-5. Five enzymes were
identified as being involved in
this degradation process: LigD/L
(Cα-dehydrogenases), LigF/E
(β-etherases) and LigG (glutathione
lyase). The activity of these
recombinant enzymes obtained
in our laboratory “The Protein
Factory” has been tested on the
lignin dimeric model compound
1-(4-hydroxy-3-methoxyphenyl)-
2-(2-methoxyphenoxy)-1,3-
propanediol as substrate, proving
that this enzymatic system is able
to efficiently cleave and process
this type of chemical bond in the
+
presence of NAD and glutathione.
Kinetic studies on the model
compound have been performed
and the whole process has been
optimized.
In parallel to this project, a tool
box of enzymes composed
of many different classes of
enzymes aiming to perform
selective transformation of lignin
functionalities has been built
including commercial enzymes,
recombinant enzymes and
enzymes obtained through a
metagenomics approach. The
preliminary unsuccessful results
obtained when the unmodified
raw material (ProtobindTM1000
provided by the technical manager
of the Valor Plus Project) has been
incubated with different enzymes
indicated that the enzymatic
biotransformations required a
more homogeneous starting
material.
The development of a pre-
treatment strategy became
necessary (Part B) to separate the
small molecules from the bulky
material before carrying out other
tests. The first approach has been
the dialysis of raw lignin in order
to obtain two different fractions,
a retentate rich in high molecular
weight molecules, and a permeate
enriched in small compounds. A
further step towards a physical
separation in terms of molecular
weight has been performed
in collaboration with Biobasic
Environnement (Saint-Beauzire,
France), an industrial partner
of the European Project Valor
Plus. Two different approaches
have been developed. The first
method involved a microfiltration
followed by two ultrafiltration
steps on a starting solution of
ProtobindTM1000 in ethanol/basic
water. The process has been
set-up with two different sets
of membranes (membrane in
polyethersulfone with a cut-off of
3 kDa and 1 kDa and membranes
in stabilized cellulose with a cut-off
of 5 kDa and 2 kDa). The second
method consisted of a soxhlet
extraction of ProtobindTM1000
in methyl ethyl ketone
(2-butanone, MEK) followed by two
ultrafiltration steps using stabilized
cellulose based membranes with
a cut-off of 5 kDa and 2 kDa. This
second method has also been
applied to another industrial lignin
(Kraft). All the different fractions
obtained from these sequential
ultra-filtrations had been fully
characterized by GPC, GC-MS,
ESI-MS, DSC, TGA, GFC and HPLC.
Unfortunately, when these more
homogeneous fractions (Part C)
have been incubated directly with
different enzymes, no degradation
has been identified. Probably, this
was because these polymers were
too bulky to be efficiently attacked
and modified by the protein
activity, and therefore a new
approach has been developed
based on a two-step cascade
process. The rationale behind this
approach was that a first oxidation
step (which could be chemical or
photochemical) was required to
partially cleave and oxidize the
polymeric skeleton in order to
make it more accessible to the
enzymatic activity.
Therefore, the optimization of
the parameters concerning the
chemical degradation approaches
were performed, arriving to the
conclusion that the best results
have been obtained when the
samples have been treated
with NaBr, NaClO, and TEMPO.
However, the chemical approach
is always linked to the use of harsh
conditions, solvents to extract
the products and expensive
catalysts. On the other hand the
use of a photochemical path
is very appealing as it would
be a greener alternative since
it relies on UV light to catalyze
the depolymerization. The
photocatalytic step therefore
exploits the oxidation capability of
UV-irradiated ZnO nanoparticles
in order to establish a water based
greener oxidative pathway. The
photochemical degradation tests
provided quite promising results,
as the GPC analysis indicated
a clear reduction of both Mn
(number-average molecular
weight) and PDI (polydispersity
index). These partially degraded
PhD Yearbook | 2018
317
products obtained both from the
Industrial Chemistry and Chemical Engineering
first photochemical and chemical
oxidations have then been
tested as substrates for different
enzymatic bioconversions.
Different commercial laccases:
LAC 3 from Trametes spec., LAC C
from Trametes versicolor and LAC F
from Funalia trogii provided by ASA
Spezialenzyme partner of the Valor
Plus Project have been assayed.
The time course experiments
performed on all three enzymes
highlighted the best operative
conditions for the bioconversions.
The cascade process indeed leads
to a clear reduction of the Mn
and Mw values, even if the total
yields in terms of mass recovery
of the two-step process on 1 g
scale in each case does not exceed
20 %. Further studies are now
in progress for the optimization
of the whole process for the
application in a pilot plant.
 318
AG2STM (Acid Gas to Syngas) Technology
Andrea Bassani - Supervisor: Prof. Flavio Manenti
CO2 is dramatically affecting
our life: climate change and
consequent social-economic
problems are identified as Grand
Challenges of current century.
Indeed, it is estimated that average
overall CO2 emissions were 35 Gt/y
in the last years. Unfortunately,
due to its thermodynamic
stability and low chemical value,
CO2 presents minor industrial
uses. One of its most important
consumptions in chemical plants
is in urea synthesis. Catalytic dry
reforming or hydrogenation of CO2
represent a viable application for
CO2 consumption; nevertheless,
these have not been successfully
applied to industrial scale yet,
because of considerable energy
consumption and need for large
amounts of hydrogen. Direct
production of methanol from
CO2 is also a potential route, but
catalysts and reactors are still
in a research phase. However, a
new potential way to reuse CO2 is
available: AG2STM technology. The
latter allows to convert CO2 into
syngas by means an oxy-reduction
reaction with H2S which takes
place into a regenerative thermal
reactor (RTR):
Moreover, AG2STM technology
allows to convert another
critical molecule: H2S. Indeed,
H2S is a poison for industrial
and automotive catalysts and a
source of corrosion for process
equipment. The Claus process
is the most widespread large-
scale technology adopted for the
neutralization of H2S. Its thermal
stage promotes the conversion
of H2S into elemental sulfur and
water by a controlled oxidation
at high temperatures (900-1500
°C). In this case, H2S and CO2
are reacted and a mixture rich
in H2 and CO (known as syngas)
is obtained with water and
elemental sulfur as harmless
by-products. The scope of the
Ph.D. activity is to propose and
demonstrate the industrial
feasibility of this innovative and
sustainable process. As first,
the involved kinetic mechanism
was studied in deep, starting
from the simplest mechanisms
Fig. 1 - PFD of AG2STM Technology
of H2S (i.e. H/S system) pyrolysis
up H/S/O/C global system. The
kinetic was extensive validated
with literature and industrial cases
and in the well-known conditions
of Claus Process. This activity was
complemented by the integration
of the kinetic mechanisms at the
process-scale for studying the
industrial feasibility of the process.
Indeed, coupling two different
software, i.e. Aspen HYSYS and
MATLAB®, a complete plant
model, able to manage the recycle
of unconverted acid gases has
been developed (Figure 1).
The core of the process are the
RTR where main reaction takes
place at temperatures higher
than 1000 °C. High temperatures
are necessary to (i) activate the
reactive system from chemical-
thermodynamics standpoint;
(ii) quicken kinetics; and (iii)
obtain high yields. According to
Figure 2, the RTR is composed
of two sections, where CO2/
H2S reaction (regenerative
section) and enthalpy recovery
(recovering section) are carried-
out. Due to high temperatures, the
regenerative section corresponds
to a refractory lined chamber; the
reaction is assisted by a minor
injection of oxygen for energy self-
sustainability. Enthalpy content
of RTR effluents, comprised of
products and unreacted feed
species, is recovered in a heat
exchangers train. This train, a
waste heat boiler and a feed/
effluents exchanger, is part
of the RTR as it contributes to
overall yield by quenching of
thermal reactions and energy
self-sustainability.
At RTR outlet, effluents are
subject to a further cooling in
condenser heat exchanger, where
by-products (elemental sulfur and
water) are removed. Effluents
are subsequently processed in
catalytic reactors for converting (i)
residual SO2 into elemental sulfur
(Claus reaction) and (ii) traces
of sulfur-compounds (COS and
CS2) into H2S and CO2 (hydrolysis
reaction). Finally, the main stream
undergoes a separation treatment
where unreacted species and H2/
CO mixture are split (chemical
washing); resulting streams are
recycled to RTR and exported to
the next process, respectively. The
catalytic reactors are simulated
with global reactions (Claus
reaction and COS/CS2 hydrolysis).
In order to account for the
PhD Yearbook | 2018
2016
319
Industrial Chemistry and Chemical Engineering
Fig. 2 - Regenerative Thermal Reactor (RTR) configuration
distribution of sulfur allotropes, of organo-sulfur compound during
the thermodynamic equilibrium coal gasification, was developed
is assumed. Moving from the in order to able to predict H2S
existing plant ENI® Claus plant in formation. As result for coal to
Livorno refinery, an evaluation of methanol process was found a
revamping technical feasibility of reduction of 1% of CO2 emissions
AG2STM technology is analysed. in parallel with an extra production
Simulations of the standard and of 0.5% of methanol using the
of the revamped Claus plants same amount of inlet feedstock.
are compared. The advantages
of the revamped process are
highlighted, both at technical
and environmental levels.
Finally, a preliminary economical
assessment is performed to
evaluate the Pay Back Time
(PBT) of the Tecnimont KT Claus
process. By varying the syngas
price, a variable PBT of 16-36
months is obtained (excluding
the two years of installation
without any revenues). This result
is acceptable from an industrial
point of view. Finally, AG2STM
technology is applied to steam
reforming and to CTM (coal to
methanol) process. Focusing on
the second one, a kinetic based
gasifier model (GASDS tool) was
validated and implemented in
the global process. Moreover, in
cooperation with Sotacarbo S.p.A.,
a gas-solid kinetic scheme, that
allows to study the devolatilization
 320
ADDITIVE MANUFACTURING AND METALLIZATION
FOR 3D MICROFABRICATION OF FUNCTIONAL
DEVICES
Roberto Bernasconi - Supervisor: Prof. Luca Magagnin
Tutor: Prof. Marinella Levi
3D printing is one of the
technologies that are having the
greatest impact on contemporary
world. In fact, the introduction
of additive manufacturing
techniques is slowly changing
the way industrial production is
conventionally intended. One
of the fields that may be most
affected by the introduction of
these manufacturing techniques
is microfabrication, since a
considerable number of modern
products contains or will contain
microfabricated parts, e.g. sensors,
microactuators and many others.
For this reason, exploiting 3D
printing for the so-called additive
micromanufacturing is one of the
most intriguing research topics in
modern industrial engineering.
3D printing techniques suitable
for microfabrication are however
characterized by a common
drawback: the materials used
for printing are polymers. Since
many applications require metallic
properties, it is fundamental to
individuate a technique to impart
such properties to the surface of a
3D printed object.
In the present work, a
hybrid approach combining
stereolithography 3D printing
and wet metallization techniques
was developed and applied to
the production of functional
microdevices. Stereolithography
(SLA) was selected as the most
suitable 3D printing technique to
perform cost effective 3D printing
at the submillimetric scale. Wet
metallization, both electroless and
electrolytic, was selected to impart
metallic properties to the printed
microdevices.
Wet metallization of SLA printed
resins was investigated in the first
part of the work. The possibility
to obtain good quality metallic
coatings using electroless
deposition was demonstrated on
a set of commercially available
and self-formulated photocurable
polymers. Pretreatment for
metallization was found to have
a critical influence on the layers
obtained. Different etching
methods were tested, but only
the alkaline pretreatment proved
able to guarantee adhesion
between the metal coating from
electroless deposition and the
resins. In particular, exposure to
-
OH containing solution was found
to break the ester bond present in
most resins due to the presence of
acrylate moieties, dissolving thus
the photopolymerized material
and increasing surface roughness
(figure 1).
Optimized alkaline etching
parameters were found to be
ideal to provide acceptable
adhesion and surface quality to
electroless plated Cu and NiP
layers. Electrolytic deposition
was attempted on the surface
of 3D printed resins made
conductive by a first layer of
electroless deposited metal. Many
commercially plated metals and
alloys were deposited on Cu and
NiP base coatings. Both decorative
and functional applications were
demonstrated.
After optimizing the metallization
process and demonstrating its
feasibility for the manufacturing
of functional devices at the
microscale, two realistic
application fields were selected:
MEMS and microrobotics.
Concerning MEMS, the first
3D-printed and electroless
metallized z-axis accelerometer
(figure 2) has been designed,
fabricated and tested. The
Fig. 1 - DL260 resin surface
morphology after alkaline
pretreatment
tri-dimensionality of the
3D-printing fabrication process
has been exploited in the design
process of the mechanical
structure and the electroless
metallization has been employed
to allow the capacitive testing of
the device. A good agreement
between the performances
predicted by FEM simulation
and the experimental tests
has been found, thus showing
a very good reliability of the
proposed fabrication process.
The work demonstrated that
3D printed and wet metallized
accelerometers show comparable
or better performance in terms
of sensitivity and linearity with
respect to current state-of-the-
art accelerometers. The marked
advantage of the proposed
devices is the possibility of full
customization and the lack of
Fig. 2 - 3D-printed and electroless metallized z-axis accelerometer: visual
appearance (a) and capacitance variation vs. acceleration (b)
many design constrains typical
of the current MEMS production
process.
Regarding the microrobotics
applicative field, rolling actuated
microrobots for cell deliver
and water remediation were
produced. Both devices showed
full actuability, with a complete
control over direction and
speed of motion, together with
a demonstrated functionality.
Microscaffolds were found to
be biocompatible and able to
transport cells. They are intended
for use in zones of the human
body characterized by the
presence of cavities presenting
suitable dimensions. These may
include the gastrointestinal
system or the respiratory system.
Photoactive/antimicrobial devices
were found to be efficient in
PhD Yearbook | 2018
2016
321
Industrial Chemistry and Chemical Engineering
photodegrading pollutants and
in killing bacteria. Their main
application is the control of water
quality through the reduction of
pollutants concentrations and
the containment of bacteria
proliferation. They can be used to
improve water conditions locally
inside closed pipes, canals and
water reservoirs by positioning
them exactly where their action is
needed.
All the devices presented proved
to be fully functional, confirming
thus the feasibility of the hybrid
3D printing/wet metallization
approach for the additive
micromanufacturing of functional
devices.
Fig. 3 - Time lapse representation of
a circular actuation pattern for a 3D
printed and wet metallized scaffold
for cell delivery
 322
Fundamental investigation of the transport
properties in open-cell foams
Mauro Bracconi - Supervisor: Prof. Enrico Tronconi
Open-cell foams are widely
acknowledged as potential
enhanced substrates for catalytic
reactors, due to a number of
promising features. High void
fractions, low pressure drops, and
the intense gas-to-solid transfer
rates make foams attractive
for many different catalytic
processes in the quest for a more
environmentally and energetically
sustainable world. Moreover,
their totally interconnected solid
matrix grants effective heat
conduction through the structure,
strongly improving the global
heat transfer performances of
the support. Despite their great
potential, however, the application
of open-cell foams to industrial
catalytic processes is limited by
the poor understanding of the
related transport phenomena.
The large deviations between
the experimental data and the
predictions of the available
literature correlations for
momentum, mass and heat
transfer in foams point out the
need for a painstaking study of the
relevant transport mechanisms.
This thesis aims at the
fundamental analysis of the
transport properties of open-cell
foams by means of Computational
Fluid Dynamics (CFD). CFD
provides, in fact, a detailed insight
into the flow field and in the
complex associated transport
mechanisms. Moreover, it can
be employed for the derivation
of lumped parameters and
descriptors by means of the
hierarchical approach, thus
enabling CFD-based engineering
correlations. The application
of this approach requires new
methodologies for the generation
of reliable virtual structures, in
order to obtain computational
domains representative of the
complex geometry of foams.
Moreover, the interpretation
of the simulation data requires
adequate models for the foam
morphological parameters,
like e.g. specific surface area,
characteristic lengths.
The focus of the first part of
the work is the solution of the
main numerical and geometrical
problems. In particular, a
systematic procedure for the
virtual generation of open-cell
foams has been developed. Based
on a few pieces of readily available
geometrical information, i.e.
porosity, cell size and strut cross-
sectional shape, this methodology
provides virtual foam samples (see
Figure 1) which totally retain all
the geometrical and topological
features and properties of real
structures. Moreover, the virtual
structures accurately reproduce
the hydrodynamic behavior of real
open-cell foams. In doing so, a fully
theoretical geometrical model has
been also derived. This model is
based on a simplified description
of the foam structure by means
of an ideal tetrakaidekahedral
unit cell and is able to predict the
geometrical features of real foams.
Moreover, it provides accurate
first guesses for the virtual
reconstruction procedure.
The solution of those problems
granted the instruments required
for the accurate investigation of
the transport properties and of the
underlying mechanism by means
of CFD. In particular, the analysis
of the gas-to-solid transport
properties has been addressed.
The investigation of external mass
transfer has been performed
in combination with dedicated
experimental activity. The effect
of the morphological features has
been parametrically investigated
and the results used to derive
lumped parameters. It has been
found that the mass transfer
coefficients mainly depend on the
porosity of the foam. A decrement
in the porosity increases the mass
transfer properties of the support.
A negligible dependence on the
strut cross-section shape and on
the cell size has been obtained.
An engineering correlation able to
accurately predict the transport
properties has been derived as
shown in Figure 2.
The pressure drop has been
analyzed by means of an
innovative approach. Virtual
samples have been manufactured
by means of 3D printing. This
removes the discrepancies
between the real foams and
geometrical models and enables a
precise control of the geometrical
properties of the investigated
samples. Numerical simulations
on the virtual models and
experimental tests on the 3D
printed samples are employed
jointly to understand the effect
of the porosity, cell size and strut
shape on the pressure losses. As
expected, an increase of both the
porosity and the cell size reduces
the pressure losses, whereas
the pressure drop increases by
changing the strut shape from
circular to triangular. These results
have enabled the derivation of
an engineering correlation for
pressure drops based on an
Ergun-like approach. A detailed
comparison of open-cell foams
against other conventional
catalytic supports has been carried
out. Open-cell foams present
Fig. 1 - Final reconstructed foam samples with circular (a)
and triangular (b) struts (c)
higher mass transfer rates with performances at fixed porosity.
PhD Yearbook | 2018
2016
respect to honeycomb monoliths This result enables the derivation
at the cost of higher pressure
drop. In this perspective, the
of geometry of open-cell foams
that can increase the heat
trade-off between external mass transfer by adopting a constant
transfer and pressure losses strut diameter along the axis.
shows that the performances of These structures cannot be
foams can be comparable or even manufactured by conventional
higher than the ones of monoliths techniques, but the virtual model 323
in a limited range of operating can be easily generated by additive
Industrial Chemistry and Chemical Engineering
conditions. manufacturing techniques such as
The global heat transfer metal 3D printing.
performances of foams have Finally, I have performed a
been investigated aiming at qualitative analysis of the
establishing the impact of the interactions between multiphase
geometrical parameters on their flows and open-cell foams by
effective thermal conductivity. carrying out numerical simulations
In doing so, a parametrical aiming at evaluating the static
investigation of the effect of the holdup and the liquid spreading
geometrical properties has been through the foam structure.
performed. The solid fraction has The numerical results are in line
been identified as the controlling with experimental data present
parameter along with the solid in literature, thus assessing the
distribution between nodes capability of the CFD simulation
and struts. An increase of the in the investigation of these
solid fraction promotes the heat phenomena as well.
conduction in the structure, while
a reduction of the ratio between
node and struts increases the
Fig. 2 - Sherwood number as a function of the Reynolds
number for experiments (empty symbol) and CFD
simulation (full symbol) along with the mass transfer
correlation (full line)
 324
The combination of ROP and RAFT
Polymerization for the Synthesis of
Polymeric Nanoparticles
Umberto Capasso Palmiero - Supervisor: Prof. Davide Moscatelli
Polymeric nanoparticles (NPs)
are colloids in the nanometer
size range that find applications
in several fields, such as optics,
coating, and medicine. In this
latter case, they can be used
as drug delivery systems
for different therapeutics,
ranging from lipophilic drugs
to oligonucleotides, in order to
improve their pharmacological
profile, efficacy and to avoid
possible side effects in the
treatment of several illness, such
as cancer. These nano-colloids
typically consist of polyesters
which are able to degrade
into safe and easily removable
compounds, such as lactic acid
and hydroxycaproic acid. However,
NP production is often a complex
process that requires mixing
devices. In addition, expensive
purification steps are necessary
to eliminate the unloaded drug
and the high amount of organic
solvent used in the NP production
step. At the end of this process,
a lyophilization step is generally
adopted to assure a good shelf-life
of the final product. All the above-
mentioned steps hamper the cost-
effective use of a re-formulation
of the same therapeutic agent
and, in turn, reduce the availability
of these treatments among the
patient population. The aim of this
PhD thesis is to solve this issue
and to develop a novel single-step
NP production protocol which
requires only a syringe and a
needle, thereby eliminating the
subsequent purification and
freeze-drying steps. This advance
build on the optimization of the
hydrophilic/lipophilic ratio of
block-copolymers that are able
to directly self-assemble in water.
The additional degree of freedom
necessary for this optimization
was introduced via the synthesis
of these materials thorough the
combination of the reversible
addition-fragmentation chain
transfer (RAFT) polymerization,
a technique that allows the
control over the structure and the
molecular weight of the aliphatic
polymers, and the ring opening
polymerization (ROP), the most
diffused process adopted for the
production of polyesters. The first
part of this work was devoted to
the modeling and kinetic study
of the RAFT polymerization
and copolymerization of water
soluble monomers in order to
better understand the effects of
the reaction parameters on the
structure of the final polymer.
Subsequently a general model
of the NPs formed by a block-
copolymer synthesized via RAFT
polymerization of a hydrophilic
and hydrophobic monomer has
been developed. A method that
consists in the use of amphiphilic
RAFT macromolecules (the
so called “RAFT surfmers”)
has been found to decouple
the dependence of the main
characteristics of the NP and
polymer molecular weight. A good
agreement between the prediction
of the model and the experimental
results has been obtained. As an
example of the application of this
technology, fluorinated NPs with
different surface charges that are
useful for optical experiments
have been produced. Then the
use of RAFT macro-surfmers
has been combined with ROP to
produce a library of biodegradable
amphiphilic block-copolymers able
to self-assemble in water via the
use of only a syringe in order to
avoid all the post-process steps
necessary to store them. In fact,
the classic polymeric surfactants
developed so far require a mixing
device to form NPs and to load
the drug, a subsequent dialysis
to eliminate the large amount of
organic solvent, the lyophilization
to avoid NP degradation during
the storage and finally the
reconstitution of the colloidal
suspension before injection. Most
of these steps are also required
in the NPs obtained via emulsion
polymerization of biodegradable
vinyl monomers, as visible in
Figure 1.1. The main difference
between the two types of NPs
resides on the loading step that
needs to be carried out after
PhD Yearbook | 2018
2016
325
the production of the colloids
Industrial Chemistry and Chemical Engineering
in the case of the emulsion
polymerization. In this work,
however, the block-copolymer is
optimized to self-assemble in the
reduced turbulence conditions
of a syringe needle. In this way
it can be stored as a dry powder
and then dissolved with the
therapeutic in a small amount of
organic solvent and formulated
into NPs just before injection.
As a proof of concept, trabectedin
(ET-743), a potent anticancer
therapeutic that causes necrosis
of the blood vessel at the site of
injection, has been loaded on
this novel platform with high
efficacy. The ability of the NPs to
retain the drug inside the core
allowed to reduce the “phlebitis”
side effect without affecting
the trabectedin antitumor
activity, potentially leading to
a better patience compliance.
The previously optimized block-
Fig. 1 - Classic pathways for the synthesis of polymer NP-based therapeutics
copolymer has been also used to
load paclitaxel, another important
anticancer therapeutic. The
resulting formulation has been
found as effective as the most
diffuse commercially available
formulation (Taxol®), but without
the use of Chremophor EL, a toxic
surfactant known to cause several
adverse side effects.
 326
Catalytic Partial Oxidation of Liquid Fuels:
Experimental and Modeling Analysis of a
Lab-Scale Adiabatic Reformer
Andrea Carrera - Tutor: Prof. Carlo Cavallotti
Advisor: Prof. Gianpiero Groppi
The need of efficient processes
for energy conversion and
abatement of pollutant emissions
keeps high the interest towards
efficient energy systems, such as
fuel cells, and prompts towards
the improvement of combustion
and exhaust after-treatment
systems. Hydrogen is attractive
for its possible usage as energy
carrier as well for the reduction
of pollutant emissions. The
superior combustion properties
of hydrogen can be exploited in
Internal Combustion Engines or
Gas Turbines, for the generation
of mechanical work, electricity
or heat. Syngas, i.e. a mixture of
hydrogen and carbon monoxide,
when purified from particulate
matter and sulfur compounds,
may also be fed to Solid Oxide
Fuel Cells, for the conversion into
electric energy. The development
of SOFC-based Auxiliary Power
Units has a great interest because
of the availability of electric energy
when the engine is off. Synthesis
gas may also regenerate Lean NOX
Traps during their rich phase.
Hydrogen may be either stored
or produced on board of vehicles.
Though the stationary production
of hydrogen is well established,
the first solution should face many
challenges before becoming a
commercial technology. In fact,
hydrogen storage is costly, and
requires specific infrastructures
for its distribution. Hydrogen
on-board production represents
an interesting alternative, as
several technologies allows the
conversion of primary fuels
into this energy vector. The
development of a compact
and light device for hydrogen
production is necessary when
considering mobile applications.
Currently, the main commercial
hydrogen production processes
are steam reforming (SR),
partial oxidation (PO, either
non-catalytic, or catalytic), auto-
thermal reforming (ATR). Steam
reforming is a well consolidated
technology, which consists in the
reaction between hydrocarbons
and water, in the presence of a
catalyst. Though steam reforming
is widely used in the industry, it
is an endothermic and energy-
demanding process, which
requires external heat supply,
therefore it is not suited for
small-scale applications. In the
Partial Oxidation process, oxygen
reacts with hydrocarbons to
produce carbon dioxide, water,
carbon monoxide and hydrogen.
This technology works in harsh
conditions, hence it is not valid
for the small-scale hydrogen
production, Catalytic Partial
Oxidation requires shorter
contact times (t ~ ms) than the
non-catalytic process (t ~ s) to
achieve equilibrium. The CPO
process is globally exothermic
and this allows the use of simple
adiabatic reactors. In the catalytic
Auto-Thermal Reforming, both
water and oxygen are reacted
with the hydrocarbon feedstock.
This process may be thermally
neutral or exothermic, according
to the reactant composition
employed. ATR operates at lower
temperatures than CPO, but the
former requires slightly higher
reactor volumes, as well as water
storage and dosing systems.
The CPO process is globally
exothermic and may be carried
out in adiabatic reactors, loaded
with a structured catalyst, which
offer a better trade-off between
pressure drops and transfer
properties with respect to random
packings. Rh catalysts offer
several advantages, compared to
transition metal and Pt catalysts,
such as high activity and selectivity
to syngas, low tendency to form
coke and good thermal stability.
In this work, the CPO of iso-
octane and n-octane has been
investigated on Rh based catalysts,
supported over a 400 CPSI
cordierite honeycomb monolith in
a lab-scale auto-thermal reformer,
equipped with the spatially
resolved sampling technique. A
one dimensional, single-channel,
heterogeneous, adiabatic reactor
model have been used to verify
the reaction mechanism and to
better understand the interplay
between chemical kinetics and
transport phenomena.
Both iso-octane and n-octane
exhibit an indirect-consecutive
reaction scheme, with oxidation
of the fuel to CO2 and H2O and
reforming of the excess fuel with
H2O with production of syngas,
which results in the presence of
an Oxy-Reforming zone next to
the catalyst inlet and a Reforming
zone downstream, and leads
to the onset of a hot spot of
temperature.
The impact of mass transfer on
the CPO of octane isomers is
important, and affects significantly
the thermal management of
the reactor. In fact, due to the
decrease of molecular diffusivity
with increasing size of the
hydrocarbon molecule, and
thus to the decrease of steam
reforming efficiency, the hot spot
temperature increases when
passing from methane to octanes,
at constant carbon and oxygen
flow rates, as detailed in Figure 1.
Temperature profiles along the
reactor axial coordinate are
similar between the two octane
isomers, as reported in Figure
2. High temperatures in the gas
phase (T>700°C) activate gas
phase cracking reactions, with the
production of olefins and light
alkanes, which can condensate,
forming soot or coke, or can
further react on the catalyst
surface, producing syngas. While
the heterogeneous reactivity of
the two isomers, n-octane and iso-
octane is similar, the two octanes
undergo different gas phase
decomposition pathways, which
lead to different distributions
of olefins and light alkanes: iso-
octane is more selective towards
iso-butylene and propylene,
while n-octane mainly produces
ethylene.
Temperature Programmed
Oxidation revealed that different
forms of coke are deposited
onto the catalytic monolith.
Coke deposition mainly occurs
in the reforming zone and is
more pronounced in the case of
n-octane. Deposition of carbon
species is thermodynamically
favored by N2 dilution, which
was adopted in the experiments
to prevent catalyst overheating.
However, reactant dilution can
be practically implemented
in combination with Internal
Combustion Engines using the
Exhaust Gas Recirculation.
Thermodynamic calculations
show that, thanks to the H2O
content of the diluting gas, this
is effective in the reduction of
the operating temperature of
Fig. 1 - Comparison among
experimental temperature profiles in
the CPO of methane and iso-octane;
methane CPO: CH4=27.3%, C/O=0.9,
F=10Nl/min, TIN=25°C; iso-octane CPO:
iso-C8H18=4%, C/O=0.9, F=10Nl/min,
TIN=85°C.
the reactor, while keeping the
PhD Yearbook | 2018
2016
process in a thermodynamically
safe zone with respect to coke
deposition. Still, coke deposition
is an important issue, especially
when using alumina as support.
The introduction of a more basic
support with higher surface area
than α-alumina, such as MgAl2O4 327
spinel, contributes to enhance
Industrial Chemistry and Chemical Engineering
the rate of steam reforming, with
the twofold benefit of lowering
the hot spot temperature and
hindering the deposition of carbon
species on the catalyst; both
aspects contribute to preserve the
catalyst stability under octanes
CPO conditions. TPO analysis
confirms that the alumina support,
more acidic, is subjected to higher
coke deposition than magnesium
aluminate. The co-feed of
aromatics and naphthenic
hydrocarbons and sulphur
poisoning should be addressed in
more detail in the future.
Fig. 2 - Comparison among
experimental temperature profiles in
the CPO of iso-octane and n-octane;
x-C8H18=2%, C/O=0.9, F=10Nl/min,
TIN=85°C.
 328
Evaluation and development of biomass
conversion processes for the chemical and
biochemical valorization of lignin
Chiara Francesca Carrozza - Supervisor: Prof. Citterio Attilio
Tutor: Prof. Brenna Maria Elisabetta
Through the integration of green
chemistry into biorefineries and
the use of low environmental
impact technologies, future
sustainable production chains of
biofuels and high value chemicals
from biomass could quickly
forecast. Despite the fact that
a massive amount of lignin is
obtained from the biorefineries,
the use of lignin in real industrial
processes is far to be obtained.
Hence, approaches to develop
lignin conversion into high-value
co-products are progressively
investigated in the attempt to
improve the economics of the
biorefineries.
In this sense, the PhD project deals
with the valorization of biorefinery
lignin as by-product from second-
generation bioethanol production
involving either a biochemical or
chemical treatments. The starting
materials were obtained from
different stages of the cellulosic
ethanol plants powered by
Proesa™ technology and operated
in Crescentino (Italy).
More specifically, the research
activity was divided in four main
Chapters:
Chapter 1: Introduction
In order to face the challenges
of becoming independent of
crude oil and switching to a more
sustainable and carbon neutral
society, biomass has emerged as
most prospective raw material
in the next future. The basic
knowledge about plant origin
materials, with specific attention
to lignin component will be
introduced in this chapter. In this
sense, a deep understanding of
biomass as renewable feedstock
will be provided. Then, the
sustainable use of lignin, as the
other major constituent in biomass
and its potential as starting
material for the production of
chemicals due to its aromatic
structure, will be investigated.
Several approaches either focus
on the direct utilization of lignin
with suitable modification to find
application in low value products
including dispersant, binders,
emulsifiers and resins, or target
the depolymerization of lignin
macromolecule into high value
aromatic fine chemicals will be
presented. Finally, a general
overview of Crescentino plant and
its technology will be reported in
order to understand where the
raw materials used among all
experiments come from.
Chapter 2: Lignin isolation and
characterization
Starting from protocol procedures,
optimization of extraction and
purification treatments combining
mechanical treatment (ball milling)
and an intensive washing step will
be performed. All the materials
were fully characterized by
means of different spectroscopic
techniques and their properties
compared with the raw material. It
was found that soxhlet extraction
as washing procedure will lead to a
material containing 96% of lignin.
Meanwhile, in collaboration
with the University of Pavia, the
effect of the growth of two fungal
species, Trichoderma asperellum
EVT4 and Pleurotus ostreatus, which
respectively are able to degrade
polysaccharides or lignin aiming
the biodegradation of the material,
was tested. It was found that
Trichoderma asperellum EVT4 is
able to purify the material lowering
the content of sugar (reducing
sugars or glucose) in a strong
way but leaves a complicated
tri-dimensional structure.
Comparison between chemical
and biochemical purification have
been done and it can be assumed
that both procedure can purify
lignin rich residue with a good
degree of lignin purity but with
different structural properties.
Finally, in collaboration with
Lawrence Berkeley National
Laboratory, a complete biorefinery
process (pretreatment, hydrolysis
and fermentation) in lab scale
using ionic liquid as solvent to
degrade and solubilized two
different feedstocks, switchgrass
and poplar, was performed.
Compared with the traditional
pretreatments, IL pretreatment
is a relatively new approach and
has several parameters that need
more investigation. The main
drawbacks in ILs pretreatment
consist in the extensive washing
procedure of the pretreated
biomass to remove residual
amounts of ILs that may inhibit
downstream saccharification and
fermentation. To overcome this
issue, recently a new wash-free
approach was developed. It is
based on a one-pot system where
IL pretreatment, saccharification
and fermentation followed by
direct extraction of sugar and
recovery of lignin will take place
simultaneously. The one-pot
process performed gaves us
good result in term of ethanol
yield since the whole hydrolysed
glucose was converted. Hence,
it was demonstrated that IL
pretreatment improves the
“delignification” process and
significantly increases the
enzymatic hydrolysis efficiency.
Physical and chemical alterations
in lignin structure were
investigated with the aim to
understand its reactivity during the
whole IL treatment.
Chapter 3 and Chapter 4:
Catalytic oxidation of lignin
and lignin depolymerization by
hydrogenolysis.
Since lignin is the richest source of
renewable aromatic compounds
on earth, its conversion into
well-defined aromatic chemicals
was deeply investigated by
either oxidative and reductive
approaches in mild conditions
of temperature and pressure.
As it is known in literature, the
recalcitrance of lignin and its
ability to recondensate represent
a challenge that inhibit an efficient
depolymerization.
In the first case, two different
catalytic conditions were tested
on the purified material by using
CuSO or Co(salen) as catalysts.
4
With copper sulphate the
three main aldehydes (vanillin,
siringaldehyde and p-hydroxy
aldehyde) were produced at
pressure equal to 3 bar with an
overall conversion of about 5%
(coherent with literature results).
Instead by using Co complex,
p-benzoquinones were formed
(mainly methoxy benzoquinone
and dimethoxy benzoquinone)
with a total yield of 7.2% at 3.5 bar.
On the other hand, the purified
materials were depolymerized
by catalytic hydrogenolysis
in alcoholic media with a
combination of palladium on
PhD Yearbook | 2018
2016
329
Industrial Chemistry and Chemical Engineering
carbon and/or nickel(II) acetate as
catalyst. To increase the efficiency
of the reaction, in-situ preparation
of the nickel-based catalyst was
carried out by adsorption of
metal salt on the material. In
1,2-propanendiol as solvent at 210
°C, up to 80% of the material was
solubilized with an overall yield of
reduced monolignol derivatives of
about 18%. Depolymerized lignin
fragments (both the liquid residue
and the solid residue recovered
after precipitation with water)
were detected by MALDI-ToF
analysis with a molecular weight
between 400-1400 Da.
To this end, catalytic
depolymerization reaction through
hydrogenolysis was performed
on the one-pot ionic liquid
residues. Different conditions of
temperature and pressure were
tested: at 200 °C the alcoholic
solvent (1,2-propanendiol) acts
as source of hydrogen, while at
90 °C and 140 °C the reaction
was performed adding 20 bar of
H . Solid residues, from either
2
switchgrass or poplar, reach a
maximum value of about 35%
but the distribution of product is
opposite: about 60% of S unit in
switchgrass residue and about
60% of G unit in poplar residue.
 330
Towards engineering of solid-state
supramolecular rotors via halogen bonding
Luca Catalano - Supervisor: Prof. Pierangelo Metrangolo
My PhD research activity had a
twofold objective:
a) Studying halogen-bonding
interactions (XB) at a fundamental
level by focusing on new tools
for the identification and
characterization of this non-
covalent interaction;
b) Engineering crystalline
supramolecular rotors based on
halogen bonding and studying the
effect of the strength of XB on the
dynamics of these systems.
Far-IR spectroscopy is seldom
employed to prove the occurrence
of halogen bonding. This is
essentially due to low frequencies
(< 400 cm-1) and weak intensities
of the vibrations involving
halogens. However thanks to
new standard instrumentations
the detection limit of vibrational
spectroscopy is moving to
lower frequencies allowing the
exploration of the Far-IR region.
To proof this concept, we
built a prototypical series of
supramolecular adducts based
on halogen bonding and then
we characterized them with
X-ray diffraction, DSC and IR
spectroscopy. We coupled the
experimental IR analysis with
calculations in order to correctly
assign the different vibrations to
the peaks of the IR spectra. The
adducts were formed between
strong halogen bond donors
(iodopentafluorobenzene and
bromopentafluorobenzene)
and strong Lewis bases (pyridyl
moieties and TMEDA).
All new species were obtained via
cocrystallization from standard
solvents and their crystal
structures were either confirmed
(if already known) or solved
through XRD analysis. One species
was liquid at room temperature,
thus, to obtain its crystal structure,
we successfully employed the in
situ cryo-crystallization technique.
Far-IR analysis showed red-
shifting and intensity increase of
the main vibrations involving the
halogen atoms. These effects are
diagnostic of the occurrence and
the relative strength of the XB.
Solid-state nuclear magnetic
resonance (SSNMR) spectroscopy
is a versatile characterization
technique that can provide
a plethora of information
complementary to single crystal
X-ray diffraction analysis. In order
to apply this powerful technique
to the study of XB, we designed an
experimental and computational
investigation of the relationship
between the geometry of XB and
the SSNMR chemical shifts of the
non-quadrupolar nuclei either
directly involved in the interaction
(15N) or covalently bonded to the
halogen atom (13C).
To reach this goal we prepared
two series of X-bonded cocrystals
based upon two different
dipyridyl modules, and several
halobenzenes and diiodoalkanes,
as XB-donors. All cocrystals
were fully characterized via IR
spectroscopy, thermal analysis,
single crystal and powder XRD.
We then investigated our systems
via SSNMR. For the first time, the
change in the 15N SSNMR chemical
shifts upon XB formation is
shown to experimentally correlate
with the normalized distance
parameter of the XB. The same
overall trend is confirmed by
density functional theory (DFT)
calculations of the chemical
shifts. 13C NQS experiments show
a positive, linear correlation
between the chemical shifts and
the C-I elongation, which is an
indirect probe of the strength of
the XB. These correlations can be
of general utility to estimate the
geometry of the XB.
Amphydynamic crystals are
materials built to possess rapidly
moving components in the solid
state. Nowadays there is a growing
interest in this class of compounds
for the development of new
functional materials and molecular
machines.
The aim of our research is to take
advantage of crystal engineering
principles to assemble stators
and rotators into cocrystals
shown highly efficient molecular
dynamics. This strategy has few
advantages such as its intrinsic
flexibility and hence the trivial
access to a vast number of
different supramolecular rotors.
Firstly, we were able to synthesized
crystalline supramolecular
rotors self-assembled by XB of
diazabicylo[2.2.2]octane (dabco),
a well known C3-symmetric
cylindrically shaped rotator, and
a set of five fluorine-substituted
iodobenzenes, acting as strong
halogen bond donors, that
take the role of the stators. We
characterized the adducts via
single crystal and powders XRD,
IR spectroscopy and melting
point measurements to confirm
their formation. Then we used
variable temperature 1H T1 spin-
lattice relaxation measurements
to characterize the dynamics of
the rotors. All structures display
ultrafast Brownian rotation around
the XB with activation energies
ranging from 2.4 to 4.9 kcal/mol,.
In all cases the activation energies
are lower than the one of pure
dabco (8.3 kcal/mol). Within the
cocrystals, the rotators are partially
isolated from their neighbouring
rotator molecules by XB donors,
which work both as stators and
bearings. The pre-exponential
factors of the dynamics are in the
range 1–9 x 1012 s-1. These results
are comparable to those found
in rotors with covalent rotational
axis, suggesting that XB is robust
enough to work as an efficient
axle of rotation. Lineshape
analysis of quadrupolar echo 2H
NMR measurements in selected
samples indicated rotational
trajectories consistent with both
3-fold and 6-fold symmetric
potential of the rotator.
The second step of the
supramolecular rotors project was
to understand how the strength
of XB, working as axle of rotation,
influences the dynamics of the
rotors. To do so, we successfully
realized two new halogen-bonded
supramolecular rotors with
isomorphic crystal structures at
100 K and room temperature. The
only difference lies in the halogen
atom involved in the XB (X = Br, I).
Solvent assisted mechanochemical
crystallization allowed us to
prepare pure cocrystals. The
increasing dynamic disorder
of dabco from low to high
temperature observed by single
crystal X-ray diffraction and
quadrupolar-echo 2H NMR led us
to conclude that the rotation of
dabco in both cocrystals can be
described by a six-fold rotation
potential surface with three low
energy minima and three minima
of higher energy that, by this
reason, at low temperatures are
less populated. 1H T1 relaxometry
experiments allowed us to
calculate the activation parameters
for the rotation of dabco in both
cocrystals. We observed good
correlation between the low
crystal packing, Ck = 0.66 at 100 K,
and the small rotational barriers
of dabco, 1.15 and 0.71 kcal/
mol, respectively, with the later
being the lowest reported in the
field of molecular rotors. The two
cocrystals showed pre-exponential
factors of almost the same
magnitude (τ0-1), 1.3 x 1012 s-2 and
1.3 x 1012 s-2 respectively.
Using isomorphous cocrystals, as
demonstrated by X-ray diffraction
and IR data, we were able to sort
out the origin of enthalpy and
entropy of activation. Accordingly,
PhD Yearbook | 2018
2016
by fitting 1H T1 data to the
Arrhenius, Eyring and Kubo-Tomita
equations, it was found negative
entropies of activation of the
≠
same magnitude (∆S = -3.0 cal/
mol K) supporting the similarity
of the normal modes and lattice
phonons. On the other hand, 331
the enthalpy of activation for the
Industrial Chemistry and Chemical Engineering
rotation of dabco in co-crystal
≠
involving iodine (∆H = 0.95 kcal/
mol) was shown to be almost
twice as that of the one involving
≠
bromine (∆H = 0.54 kcal/mol)
that correlates well with the subtle
crystal structure differences,
namely a more hindered
environment in the dabco cavity in
the iodine-based cocrystal.
In conclusion relevant results were
obtained in both the research
strands of my PhD career.
• We demonstrated that Far-IR
spectroscopy and SSNMR are
powerful diagnostic tools for
the detection XB and for the
characterization of its strength
and geometry.
• We synthesized and fully
characterized the first
examples of XB-based
amphidynamic cocrystals.
Furthermore we were able to
sort out the origin of enthalpy
and entropy of activation
of the rotators dynamics.
The discovery of these new
crystalline multicomponent
systems thus opens up new
avenues in the development
of new smart materials and
molecular machines.
 332
A STRUCTURAL APPROACH TOWARDS CELL
ADHESION MODULATION AND PROTEIN
ENGINEERING
Andrea Dalle Vedove - Supervisor: Dr. Emilio Parisini
Protein X-ray crystallography is a
very powerful tool that provides,
at the molecular level, the high
quality and high resolution
structural information that is
necessary for structure-function
correlation analysis, protein
engineering, structure based drug
design and molecular dynamics
(MD) simulations.
In this thesis, this technique has
been extensively used for the
investigation of the structure,
properties and possible
application of the extracellular
portion of some selected cell-
adhesion proteins of the type I
classical cadherin family and of
a deglycating enzyme from the
large Fructosyl Amino Oxidase
family, called Amadoriase I.
Cadherins are calcium-dependent
trans-membrane proteins that
comprise three clearly discernible
regions: intracellular, trans-
membrane and extracellular. The
extracellular portion is formed
by a variable number of so-called
extracellular cadherin (EC)
repeat domains, each formed by
approximately 110 amino acidic
residues and rigidified by the
presence of Ca2+ ions between
them.
They are of critical importance for
cell-cell adhesion, a fundamental
process that results in cellular
organization and tissue
differentiation, thus allowing the
formation of tissues and organs
and ultimately the development
of complex multicellular
organisms.
Other than their adhesive
function, cadherins also perform
a cell-cell signaling function. It
is well known that variations in
cadherin natural expression level
and changes in their ability to
form either homophilic dimers
or bind to selected protein
substrates correlate with the
onset and the progression of
diseases such as cancer, asthma
and chronic inflammation states.
Although the cadherin main
functions (adhesion and
signaling) have been quite
extensively investigated over
the last two decades, the
mechanism by which such tasks
are performed still needs to
be fully elucidated. Over time,
a combination of biophysical
techniques (mainly X-ray
diffraction, NMR and SAXS)
have provided a clear, albeit
still incomplete, picture of the
adhesion mechanism, leaving
the complete trajectory that
leads, very dynamically, from the
monomer to the dimeric adhesive
state and back still partially
elusive. The implications of
such lack of complete structural
characterization are important:
to date, no clear molecular bases
for cadherin homo-selectivity and
for their energy activation profiles
have been unambiguously
identified.
In this thesis work, I have
studied selected cadherin family
members at the molecular
level by single crystal X-ray
crystallography. Overall, my
goal was to contribute to the
characterization of their adhesion
mechanism. Furthermore, in the
context of this thesis, I combined
this structural analysis with
biophysical, computational and
functional studies aimed at the
development or the identification
of molecules that are capable
of modulating cadherin
homophilic adhesion. Finally,
starting from the structural
information, I have engineered
a cadherin family member
with the goal of producing
functionalized biomaterials. In
fact, due to the important role
of the protein in tissue sorting
and formation, such cadherin-
functionalized hybrid materials
may be employed as scaffolds
in tissue engineering and tissue
regeneration or be used for the
development of cell-sorting or
sensing lab-on-chip platforms.
The second project that I focused
on in my thesis is the X-ray crystal
structure of the deglycating
enzyme Amadoriase I. A member
of the Fructosyl Amino Oxidase
family, this enzyme is capable of
hydrolyzing the bond between
the amine and the sugar moiety
of a glycated amino acid.
From a biomedical point of
view, the study of this enzyme
is very important since protein
glycation reactions occur
spontaneously in the body
over time due to the sugar
present in blood. As a result
of this spontaneous glycation
reaction (usually referred to as
the Maillard reaction) proteins
are progressively glycated and,
especially those with long half-
life, tend to become heavily
crosslinked over time. These non
specific modifications negatively
affect the function of the proteins
and may eventually lead to
the development of diseases
such as Alzheimer’s disease,
arteriosclerosis, nephropathy
and retinopathy, with a higher
incidence in elderly people and
in people with abnormally high
blood sugar levels.
Moreover, deglycating
enzymes are also utilized
for the measurement of the
concentration of the glycated
form of hemoglobin (HbA1c)
in the blood. In fact, due to
the relatively long life-time
of hemoglobin (90-120 days)
and therefore its tendency to
be glycated over time, this is
a very good indicator of the
concentration of blood glucose
PhD Yearbook | 2018
2016
333
over a period of 2-3 months. for diabetes monitoring. Other
Industrial Chemistry and Chemical Engineering
Unfortunately, however, due potential applications of such
to the fact that the available molecular technology include
enzymes only work on glycated the use of such chimeric enzyme
amino acids, this process is not to slow or even reverse collagen
immediate since HbA1c must be rigidification in aging tissues as
proteolytically cleaved before the well as in the food industry, to
enzymatic detection of selected limit protein glycation in aliments
glycated amino acids can actually that need thermal treatment (e.g.
take place. UHT milk).
During my thesis, I contributed
to the structural characterization
of of the apo form and the
substrate-bound form of
Amadoriase I from Aspergillus
fumigatus. Moreover, I have
been actively involved in the first
attempts to engineer this enzyme
in order to enhance its natural
substrate recognition capabilities.
All these studies have been
conducted using a combination
of molecular biology, protein
chemistry, X-ray crystallography
and molecular dynamics
techniques. The ultimate goal
with this project is to engineer
the Amadoriase I enzyme in
order to enlarge its catalytic
cavity and allow its catalytic
activity on large substrates
such as polypeptides or even
whole glycated proteins. This
would, for instance, potentially
improve the current procedures
employed for the measurement
of HbA1c in diabetic patients,
possibly leading to new, efficient
and low cost diagnostic tools
 334

CATALYTIC CO2 VALORIZATION PRODUCING FUELS catalytic systems need promoters, experiments. By cofeeding CO at obtained material showed very

PhD Yearbook | 2018

AND CHEMICALS
Leonardo Falbo - Supervisor: Prof. Luca Lietti
The increase of the atmospheric
based Fischer-Tropsch catalysts.
intermediates, hence decreasing
both structural and electronic, in
order to boost process yields. In
this regard, the promotion with
potassium and zinc, more than
with manganese, increases the
catalyst activity and selectivity
to high molecular weight
hydrocarbons, by changing the
basicity of the catalyst surface.
low temperature, the surface CO
coverage is high and it increases
with time on stream. In this
way, the methanation becomes
kinetically inhibited due to the
lack of available hydrogen on the
metal sites. Additionally, in the
presence of CO, some adsorbed
species, identified as carboxylates,
2016
good performance in the reaction
by operating at mild process
conditions.
In conclusion, Fe- and Ru-based
catalysts can be successfully
used to obtain high added-value
products from CO2 hydrogenation,
even in the presence of H2 335
deficient feed. The results

Industrial Chemistry and Chemical Engineering

carbon dioxide concentration
is blamed as one of the main
causes of the climate change
(global warming). For this reason,
during the last 20 years several
studies have been carried out to
intensify the technologies aiming
at its reduction. The decrease
of CO2 emissions through the
increase of energetic efficiency
of the existing fossil fuel based
industrial processes and the
utilization of renewable energy
sources are obviously options of
choice. However, strategies able
to accomplish a net reduction
of the CO2 concentration in
the atmosphere are also being
considered. Among these, the
CO2 Capture and Utilization
technologies, consisting in the
chemical conversion of CO2 into
valuable products, are reaching
great industrial and environmental
appeal. In this regard, the purpose
of this research work is to gain
more insights in the intensification
of sustainable CO2 utilization
catalytic processes producing
synthetic fuels (particularly liquid
hydrocarbons and synthetic
natural gas) and chemicals
(particularly “lower olefins” in the
C2-C4 range).
The production of liquid
hydrocarbons through the one-pot
catalytic CO2 hydrogenation is
explored on typical iron and cobalt
Co-based catalysts lead principally
to the synthesis of light saturated
hydrocarbons, whereas Fe-based
catalysts produce more heavy
hydrocarbons.
On both Co- and Fe-based
catalysts, CO and CO2
hydrogenation processes follow
a common reaction pathway,
with CO acting as intermediate
in the case of CO2 conversion.
The different selectivity of the
two processes is mainly ascribed
to different H/C ratio on the
catalyst surface. During the CO2
hydrogenation, a high H/C surface
ratio is attained due to the slow
CO2 adsorption, if compared with
respect to CO adsorption. This
behavior favors the hydrogenation
reactions of the adsorbed surface
Fig. 1 - ASF plot and olefin content in hydrocarbons for Co-based (220°C, 20 barg.
H2/COx=2.5 v/v) and Fe-based catalyst (220°C, 30 barg, H2/COx=1 v/v) during CO and
CO2 hydrogenation.
the chain growth probability
and increasing the olefin
hydrogenations. Furthermore, on
iron-based catalysts, the process
yield is also controlled by the
equilibrium between the chain
growth sites (iron carbides) and
the hydrogenation sites (iron
oxides and reduced iron centers).
The former sites are not stable
in the absence of high CO partial
pressures and they can evolve
with the time on stream into the
latter sites, which are active in the
secondary reactions of olefins. The
relative abundance of these two
types of sites affects conversion
and selectivity of the catalysts.
Although the process selectivity
during CO and CO2 hydrogenation
is not dramatically different by
using Fe-based catalysts, these
The production of synthetic
natural gas is here investigated
on low-loading Ru-based catalysts
supported on alumina. The
catalyst is appropriate to carry
out the Sabatier reaction (CO2
methanation) under process
conditions relevant for the Power-
to-Gas application, with an almost
complete selectivity to methane.
The catalyst performance over a
wide range of process conditions,
previously unexplored, is
described by a developed kinetic
rate equation.
Furthermore, the reactivity
of Ru-based catalysts in the
hydrogenation of CO/CO2 gas
mixtures is investigated to assess
the possibility to carry out the
CO2 methanation even in the
presence of CO-containing feed
streams. When CO is present
in the gas feed, Ru-based
catalysts show deactivation at
low process temperature. The
extent of activity loss increases
by increasing CO/CO2 ratio in the
inlet stream and by decreasing
the catalyst bed temperature,
so by working in condition
where the CO adsorption is
enhanced. Furthermore, the
role of CO in CO2 methanation
is also investigated by looking
at the catalyst surface through
dedicated infrared spectroscopic
are formed on the catalyst surface, obtained in this work pave the
blocking the CO2 adsorption when way for the development of more
process conditions are switched efficient catalytic processes for CO2
back to pure CO2 methanation. In hydrogenation, which associated
this regard, high temperatures are with a supply of renewable H2
needed to avoid a gradual catalyst and process intensification can
deactivation due to the strong mitigate the dramatic effects of
adsorption of CO on the Ru sites elevated CO2 concentration in the
and to the presence of surface atmosphere.
blocking adsorbed species.
Eventually, the hydrogenation
of CO2 to lower olefins is
investigated by using a bulk
K-promoted Fe-based catalyst
with an iron spinel structure and
high surface area. The catalyst
preparation method and the
thermal treatments have dramatic
effects on the morphological
properties and on the crystal
structure of the final catalyst. The
Fig. 2 - Effect of temperature and pressure on CO2 conversion and CH4 selectivity.
Experimental CO2 conversion data are compared to model predictions. (Central
process conditions: 310°C, 1 ata, H2/CO2=4 v/v, 10 mol.% of N2, 5 L(STP)/h/gcat
 336
Combined advanced structural tools
for characterization of polysaccharidic
polymers
Monica Ferro - Advisor: Prof. Andrea Mele
The work object of this doctoral
thesis been focused on the study
of different polysaccharidic
materials with different
characteristics. Linear and
branched polysaccharides of both
natural and synthetic origin have
been studied in solution, in the
solid or in the gel state. The work
of my PhD can be divided into two
parts: the first part (A) regarded
the study of cellulose and cellulose
derivatives; the second part (B)
was focused on the diffusion
study of small molecules loaded
in different hydrogels composed
by cross-linked polysaccharides
(Agar-carbomer) or nanoporous
cyclodextrin-based cross-
linked materials (cyclodextrin
nanosponges). Part (A). Cellulose
is the most important natural
polymer: it is mainly used in the
production of paper, but also
for the production of relevant
industrial derivative such as
cellophane and rayon. In the
last decades cellulose has also
been studied for the production
of biofuels as non-fossil fuel
source. From the analytical point
of view, cellulose was studied
by different academic and
industrial groups using different
solid state techniques such as:
Powder X Ray Diffraction (PXRD),
Fourier Transform Infra-Red
Spectroscopy (FT-IR), solid state
NMR spectroscopy, Scanning
Electron Microscopy (SEM) and
Transmission Electron Microscopy
(TEM). In my work, I mainly used
powder X-ray diffraction (PXRD)
to study cellulosic materials.
The objective was to develop
innovative, easy to apply and
accurate analytical methods for
cellulose characterization and
to monitor cellulose reactivity.
I developed an accurate
deconvolution method based on
PXRD peaks and I applied it on
industrial cellulose samples.
Moreover, I applied multivariate
statistical analysis to PXRD
data in order to evaluate and
predict the reactivity of cellulose.
Within the study of cellulose
and its derivatives, I studied
carboxymethylcellulose (CMC).
CMC is a linear, anionic polyether
cellulose derivative, generally
water soluble and with high
molecular weight. The industrial
synthesis of CMC is based on
a slurry process and involves
two steps: mercerization and
etherification. The Degree of
Substitution (DS) represents the
average number of carboxymethyl
groups present in each glucosidic
unit. The typical DS range of
commercial CMC is between
0.6 and 1.25. Several methods
to provide DS of CMC have
been previously reported in
the literature. Most of them are
based on chromatography, FT-IR
spectroscopy or Scanning Electron
Microscopy. The most commonly
used method is based on the HPLC
analysis of the hydrolysis products
of CMCs with strong acids. In
addition to this chromatographic
method, different promising NMR
methods for DS determination
have been previously described.
Some of these (both 1H and 13C)
are also based on the analysis
of hydrolyzed CMC. Others are
non-destructive methods. The
latter approaches are based on
quantitative 13C NMR in solution
(inverse GATED acquisition mode).
DS value is obtained by peak
integration. Severe limitations
of these methods arise from
the spectral overlaps generated
by the mixture of regioisomers.
Furthermore, the application is
restricted only to CMC solutions
with low viscosity in order to avoid
line broadening.
A novel, non-destructive method
based on 13C HR-MAS NMR
spectroscopy is proposed in this
work. The method herein reported
couples a wide applicability to
CMCs with different purification
degree, different viscosity and
different origin (cotton linters
or wood) with a non-destructive
procedure. Thus the results of the
analysis give more reliable picture
of the DS of the CMC in its original
state.
Part (B). In the last decades
PhD Yearbook | 2018
2016
337
hydrogels gained increasing of the pulsed field gradient
Industrial Chemistry and Chemical Engineering
interest as drug delivery scaffolds spin-echo (PFGSE) method. The
and permeation membranes experiments were repeated at
due to their ability to control several concentrations of the drug
solute transport. The diffusion in molecules dissolved in hydrogels
polymers or in porous materials having different mesh size. The
was previously studied by different signal decay provided the mean
groups using either NMR in square displacement (MSD) values
solution or solid state techniques. for each drug concentration.
However, a direct characterization The MSD along the z reference
of the active pharmaceutical axis <z2> is proportional to the
ingredient (API) in the molecular observation time elevated to
environment of the drug delivery a power factor α. The α values
formulation (i.e. in the hydrogel) define the normal Fickian random
was missing in the literature. motion (α = 1), or anomalous, non-
The innovation of the work Fickian (α < 1, subdiffusive regime,
herein discussed is in the direct α > 1 superdiffusive regime).
observation of the dynamic
behaviour of small molecules
confined in hydrogel matrixes.
The study was carried out by using
a High Resolution Magic Angle
Spinning (HR-MAS) NMR probe.
HR-MAS NMR allows to study the
molecular environment and the
transport properties of the drugs
entrapped in the hydrogels. To the
best of my knowledge, this type
of study is unprecedented in the
literature. Diffusion experiments
were performed on Ethosuximide
and Ibuprofen, taken as model
drugs, loaded in hydrogels
obtained from Agar-Carbomer (AC)
and cyclodextrins nanosponges
(CDNS). The signal intensity decays
were collected using different
observation time values (td) in
the milliseconds range by means
 338
NOVEL COBALT-BASED CATALYSTS FOR THE
INTENSIFICATION OF THE FISCHER-TROPSCH
SYNTHESIS
Laura Fratalocchi - Supervisor: Prof. Luca Lietti
The development of the optimal
cobalt based FT catalyst, with
improved activity and selectivity
to desired products, has been
the goal of this PhD thesis work.
This goal has been pursued by
a combination of experimental
activities at different scales,
starting from the preparation
of the catalysts, to their
characterization and eventually
the assessment of their catalytic
performances in the FTS.
In the first part of this thesis
work the development of a
novel preparation procedure
for an eggshell-type Co/γ-Al2O3
catalyst is reported. The main
characteristics of the obtained
material are a diameter smaller
than which is the most common
value in the scientific literature,
and a diffusive length ensuring
the absence of mass transfer
limitations. Interestingly, the
prepared eggshell catalyst shows
great catalytic performances
if compared with those of a
powdered catalyst (obtained by
grinding and sieving the eggshell
catalyst) working in a kinetically
controlled regime. This result has
been explained by considering
the onset of weak mass transport
limitations in the case of the
eggshell catalyst, which boost
both the CO conversion kinetics
and the catalyst hydrogenating
ability. These results show that
engineered eggshell catalysts with
an optimal thickness of the active
shell represent an optimal solution
for applications purpose. In fact,
it is possible to achieve high C15+
yields per mass of cobalt in the
reactor, while limiting the ∆P in
compact packed-bed reactors for
the FTS.
The second part of this thesis
work focuses in developing a
highly active Co/γ-Al2O3 catalyst
as uniformly impregnated
with a pellet size of 300 μm in
diameter. In compact FT reactors,
such diameter represents a
compromise to keep under control
mass transfer and pressure
drop issues. In this regard, two
strategies have been adopted
in this thesis work to increase
the number of cobalt metal
surface sites available for the
FTS. The decrease of the cobalt
metal particle size through the
improvement of the preparation
methods commonly applied at
the industrial scale (in view of
the necessity not to complicate
the catalyst synthesis) and the
enhancement of the catalyst
reducibility with the addition of a
small amount of noble metals (in
view of the catalyst cost).
Concerning the first approach, a
highly active Co-based catalyst has
been synthetized by diluting the
Co-nitrates impregnating solution
with an organic compound
(i.e. diethylen glycol, DEG). It
has been found that thanks to
the occurrence of combustion
phenomena between Co
nitrates, which act as oxidizers,
and DEG, which act as fuel, the
decomposition from Co nitrates
to Co oxides species during
calcination becomes exothermic
and fast. This generates highly
dispersed Co oxides particles,
which, however, are difficult to
reduce by a standard reduction
treatment in H2 at 400°C. Despite
the lower reducibility, when
compared with a catalyst with the
same formulation but prepared
without using DEG, the catalyst
prepared with DEG shows very
interesting catalytic performances,
thanks to the fact that the small
Co0 particles are intrinsically very
active. In line with the decrease of
the Co0 crystallite size, the catalyst
prepared with DEG shows slightly
higher hydrogenating activity.
Nevertheless, the increase of the
CO conversion overcompensates
the decrease of the C5+ selectivity,
thus resulting in a raised C5+ yield.
Concerning the second approach,
the possibility to increase the
number of active Co0 centers
through the addition of small
amount of platinum in the catalyst
formulation has been investigated.
In particular, the effect of
0.1wt.% of Pt has been studied
on the properties and catalytic
performances of a catalyst
prepared with (small crystallites)
and without (big crystallites)
DEG, in the Co-impregnating
solution. The Pt-promoted
catalysts have been prepared
by varying the impregnation
order of Pt and Co: (i) Pt after
Co (sequential deposition
order, SDO) and (ii) Pt before Co
(reverse sequential deposition
order, RSDO). Regardless of
the initial Co3O4 crystallites size,
all the Pt-promoted catalysts
show a strong enhancement
of the catalyst reducibility and
Co0 dispersion with respect to
the corresponding unpromoted
catalysts. This indicates the ability
of Pt in favoring the reduction
of cobalt oxides species, even
if present in small quantity.
Moreover, it has been found that
the catalytic activity trend reflects
that of the Co0 dispersion with
the catalysts prepared without
DEG, thus justifying why the
RSDO catalyst is more active than
SDO, which in turn is more active
than the unpromoted sample.
Furthermore, the selectivity to
the main FTS products is almost
unvaried with respect to the
unpromoted sample, thanks to
the small amount of Pt used. On
the contrary, the promising results
obtained with the Pt-promoted
catalysts prepared with DEG in
terms of catalyst characterization
(good reducibility and increase
of the number of Co0 sites) do
not reflect the catalyst activity,
which results (in the bad case)
almost halved with respect to the
unpromoted catalyst. This result
has been explained with the fact
that the intrinsic activity of the
Co0 sites (turnover frequency,
TOF) of the Pt-promoted catalysts
is significantly decreased. This
has been attributed to the fact
that these catalysts have a Co0
particle size distribution shifted
to lower values than that of the
unpromoted sample, thus entering
in the TOF-size range where the
FTS has been reported to be
structure sensitive.
In the final part of this thesis
work, the effect of water on
the catalytic performances of
Co-based catalyst is investigated.
This is of interest in view of the
fact that commercial FTS practices
require that Co-based catalysts
withstand long-term use at high
CO conversion, and hence at high
water concentration levels. The
obtained data indicate that water
addition leads to a remarkable
catalyst deactivation already at
low water concentration. This
phenomenon is a rather slow
process, whose rate depends on
the feed concentration of water
and whose extent depends on the
duration of water co-feeding. The
presence of water in the feed also
PhD Yearbook | 2018
2016
339
affects the process selectivity, and
Industrial Chemistry and Chemical Engineering
both reversible and irreversible
effects have been observed. In
particular, the exposure of the
catalyst to water results in an
irreversible increase of the olefin
to paraffin ratio (mostly due to
the increased olefin selectivity), of
the 1-olefins in the alkenes pool
(due to the decreased double
bond shift activity) and of the CO2
selectivity (due to increased WGS
activity). These effects have been
explained considering that the
active sites evolve during the water
co-feeding in those cobalt oxides
species which are formed as a
result of the oxidation of a fraction
of the metallic cobalt sites initially
present on the catalyst. Reversible
effects, which mostly consist
in the increased chain growth
probability and CO2 selectivity
and in the decreased CH4 and
alcohol selectivities, have been
explained by assuming that the
water inhibits the hydrogenation
reactions and acts as reactant in
the WGS reaction.
 340
Giancarlo Gentile - Relatore: Tiziano Faravelli
The thermochemical conversion of
biomass is a promising technology
to produce fungible substitutes
for petroleum-derived fuels
and chemicals. This is a topic of
increasing interest within the
scientific community in light of
the recent concerns regarding
the energy availability and the
sustainability of the industrial
processes. Thermochemical
conversion routes provide
remarkable advantages in the
exploitation of biomass because
of the possibility to handle wide
range of feedstock and address
the production of different type
of fuels and chemicals. However,
current technologies still face
significant industrial challenges,
mainly in term of efficiency
and operation scale-up, which
severely hinder the worldwide
industrialization of these
technologies. In this framework,
it requires several levels of
analysis and consideration, as
highlighted in Figure 1. Moreover,
the strong interactions between
chemical kinetics and heat/mass
transfer phenomena make the
mathematical model demanding
from a computational point of
view.
The work presented in this
thesis steps from the definition
of the methodological basis.
Firstly, a model for the biomass
characterization was presented. It
is able to predict the biochemical
composition of a generic
biomass sample starting from its
elemental analysis. The second
step regards the definition of a
comprehensive kinetic mechanism
for the biomass pyrolysis and
devolatilization. A multistep kinetic
scheme with lumped first-order

reactions was employed in this
work. The peculiarity of this
model, respect the state-of-the-
art in the literature, is a detailed
characterization of the pyrolysis
products, including a wide set of
permanent gas and tar species.
This kinetic mechanism was
eventually coupled with the
heat/mass transfer limitations
in the developed particle-
based model. The resulted
code, named bioSMOKE,
was implemented within the
®
OpenFOAM framework in
order to simulate the thermal
degradation of anisotropy biomass
particles of arbitrary shapes. The
mathematical model requires
the solution of the Navier-Stokes
equations, applied to porous
media. One of the original aspects
of this work is represented by the
development of a comprehensive
approach to model the spatial-
dependent particle shrinking
during the thermal decomposition.
It is based on the evaluation of
a shrinking wave front which
progressively evolves during the
transient and allows for handling
arbitrary multi-dimensional
particle deformations due to
anisotropy and/or non-uniform
boundary conditions. In spite of
the outstanding importance of
this topic, only few models are
available in the literature, usually
based on empirical correlations or
fitting parameters. The code was
validated against comparisons
with experimental data of
particle pyrolysis in different
range of operative conditions.
Three different geometries were
analyzed (namely, sphere, cylinder
and slab), to show the reliability
PhD Yearbook | 2018
2016
341
complex multiphase reactors (i.e.
Industrial Chemistry and Chemical Engineering
fluidized/fixed bed gasifier). In this
regard, one the critical issues is
the coupling between the particle-
scale phenomena and the reactor-
scale transport. The existing
state-of-the-art methods crudely
consider the particle degradation
because of the complexity of the
mathematical formulation and
the huge computational costs
required for tracking individual
particles. Global kinetics are
typically defined which cannot
capture the key dynamics of
particle conversion. To this end, a
reduced-order model (ROM) for
biomass particle degradation was
developed aiming to describe the
release rate of primary pyrolysis
products by means a simpler and
less demanding methodology.
 
 and the capability of bioSMOKE














spending modeling efforts in 


in handling particle shrinking.


the direction of a clearer and









 

 Despite of the unavoidable
 
 
 


deeper understanding of biomass 

 
 uncertainties of such a system,
potential is crucial towards the   the model turns out to be in
massive exploitation of this   


 reasonable agreement with the











renewable energy source.


 


 



 

 experimental data. The model
The aim of this thesis is to develop 
 





  successfully predicted the
a new comprehensive and temperature history of the particle,
predictive particle based model the mass/volume loss, as well as


 


able to provide a better insight into




 
the fundamental chemical/physical


phenomena occurring during the
biomass conversion. The proposed
problem is quite complex as
the evolution of the major species


released during the pyrolysis.
The fundamental analysis at the
particle scale is an essential step
Fig. 1 - Biomass thermal conversion: a multicomponent, multi-phase and multi-
scale towards the rational modeling of
 342
Design of liposomal and polymeric
nanocarriers capable of crossing
biological barriers
Xiaoyi Huang - Supervisor: Prof. Francesco Cellesi
This PhD thesis work focused on
the design of new liposomal and
polymeric nanocarriers capable of
crossing specific biological barriers.
Depending on the application, the
physicochemical characteristics of
the nanocarriers were tuned in order
to maximize barrier penetration and
drug delivery at the specific site.
Firstly, new targeted therapies
directed to treat glomerular diseases
in kidneys were developed, based
on the use of new protein drug-
loaded, engineered liposomes.
These lipid-based nanocarriers were
designed to target the glomerular
endothelium and release the drug
in close proximity to the glomerular
filtration barrier. Once released, the
drug should present a hydrodynamic
diameter small enough to penetrate
the glomerular barrier by diffusion
and reach podocytes, i.e. glomerular
cells which are known to play a central
role in kidney diseases. In order to
identify the key parameters which
may influence drug encapsulation and
release, different lipid formulations
and payloads were investigated.
Four thermosensitive liposomal
formulations were selected: DPPC/
DSPE-PEG2000 95/5 (mol/mol) (TSL1),
DPPC/DSPC/DSPE-PEG2000 80/15/5
(mol/mol) (TSL2), DPPC/DSPC/Chol/
DSPE-PEG2000 50/25/15/3 (mol/mol)
(TTSL) and DPPC/P-lyso-PC/DSPE-
PEG2000 90/10/4 (mol/mol) (LTSL).
Water soluble molecules, including
a low molecular weight molecule
(Carboxyfluorescein (CF)), a hydrophilic
linear polymer (dextran), a custom-
made PEGylated star polymer and a
protein (albumin) were used as model
drug molecules for encapsulation
and release. Results showed that all
liposomal formulations displayed
higher loading efficiency for
carboxyfluorescein (CF) mainly due to
its low molecular weight. The presence
of hydrophobic domains in albumin
enhanced the interaction between
this protein and the hydrophobic lipid
bilayer to some extent, thus leading
to higher encapsulation efficiency
compared to that of dextran and
PEGylated star polymer. LTSL not only
showed the highest encapsulation
efficiency for albumin compared to
other liposomal formulations, but
it was also able to release CF and
albumin in the mild-hyperthermia
temperature range (40-42°C). The
release of albumin was much more
limited than that of CF due to slower
diffusion and potential interaction with
the lipid bilayer through hydrophobic
domains. LTSL was further selected
for encapsulation and release of
therapeutic proteins, such as lysozyme
and the brain-derived neurotrophic
factor (BDNF), which has been
recognized as a potential therapeutic
agent for kidney diseases. In vitro
release tests confirmed that LTSL was
capable of retaining lysozyme and
BDNF at the physiological temperature
while releasing them under mild
hyperthermia conditions. Moreover,
BDNF maintained structural stability
and bioactivity towards damaged
podocytes during the encapsulation
process, as confirmed by SDS-PAGE
analysis and immunofluorescent
tests. The BDNF-loaded LTSL showed
good therapeutic effects towards
podocytes in standard 2D cell cultures
(Fig. 1) as well as in a 3D podocyte–
endothelial co-culture system, which
was designed to mimic the glomerular
filtration barrier in vitro. Therefore,
BDNF-loaded LTSL nanotherapeutic
may present a powerful new strategy
PhD Yearbook | 2018
2016
for effective treatment of kidney
glomerular diseases.
In order to enhance the targeting
properties of liposomes towards
glomerular endothelial cells,
specific peptides, such as
cRGD, Ac-IELLQARC-NH2 (IEL),
Ac-HITSLLSC-NH2 (HIT) and 343
Ac-CLPVASCK-NH2 (CLP), were
Industrial Chemistry and Chemical Engineering
selected as ligands and conjugated
to liposomes. Cellular uptake
experiments showed that cRGD-
conjugated liposomes exhibited
extremely high uptake by mouse
microvascular endothelial cells (EOMA)
and glomerular endothelial cells
isolated from rat glomeruli. cRGD
conjugated liposome uptake was likely
due to the overexpression of αvβ3
integrin receptors on the cell surface.
IEL conjugated liposomes also showed
substantial interaction with EOMA
due to the expression of E-selectin
(a receptor for IEL) on the tumour
vascular endothelial cells. HIT and CLP
showed enhanced selectivity towards
rat glomerular endothelial cells (Fig.
2). In vivo biodistribution of liposomes
in mice revealed that the conjugation
of cRGD to liposomes enhanced
accumulation of these nanocarriers in
the liver, spleen and kidney. Therefore,
the specificity of peptidic ligands chlorotoxin (CTX), which is known based on radiation therapy and
towards kidney should be enhanced to selectively bind to glioma cells CTX modified PNPs may have a
to improve the targeting delivery to with high affinity. The engineered great potential in enhancing blood
the glomeruli; HIT and CLP could be targeted PNPs offer the potential brain barrier penetration, and
potential candidates in achieving this for delivering therapies directly to strengthening the anti-tumor efficacy
goal. invasive brain cancer cells, thus of chemotherapeutics towards
Secondly, a glioblastoma multiform- improving the desired therapeutic glioblastoma multiform diseases (Fig.
targeted therapeutic approach based effects while minimising unwanted 3).
on the use of PLGA-PEG polymeric toxicity. Preliminary in vitro cellular Keywords: Thermosensitive
nanoparticles (PNPs) loaded with uptake and cytotoxicity studies liposomes; Polymeric nanocarriers;
the anti-tumor drug doxorubicin suggested that further modification Biological barrier; Kidney diseases;
(DOX), was developed. DOX was of the polymer structure may be Glomerular filtration barrier;
efficiently loaded into PNPs by necessary to control the release rate Glioblastoma multiform; Blood brain
nanoprecipitation, and DOX loaded of DOX and improve cell targeting. barrier; Targeted delivery
PNPs were further functionalized with Moreover, a combination strategy
 344

Hydrogen and Halogen Bonding towards α,ω-diiodoperfluoroalkanes. mesomorphic in nature, the

PhD Yearbook | 2018

Molecular Recognition and Separation
Vijith Kumar - Supervisor: Prof. Giuseppe Resnati
Molecular self-assembly has been
Halogen bonding has driven
the formation of these
anionic networks. Borromean
entanglement has been obtained
starting from all the four used
cations, all the three used
anions, but only two of the five
used diiodoperfluoroalkanes.
these molecule(s). formation of cocrystals allows for As the change of the
2016
obtained halogen-bonded
complexes exhibit monotropic
LC behaviour with smectic A
phases possibly resulting from
segregation between fluorocarbon
and hydrocarbon chains. The
obtained supramolecular
mesogens possess reactive groups 345
suitable for incorporation into

Industrial Chemistry and Chemical Engineering

well-acknowledged as one of the
efficient bottom-up approaches
to synthesize various materials
with controllable architectures and
useful properties. It has clearly
been proven that the formation of
the well-ordered structures of self-
assembled architectures can be
driven by a single interaction or by
the synergistic action of multiple
interactions. In this thesis metal
coordination, hydrogen bond (HB)
and halogen bond (XB) have been
used for the design and synthesis
of self-assembled systems tailored
to topological studies, separation
processes and obtainment
of supramolecular functional
materials.
The first part of the thesis
describes how metal coordination
enables twenty four bis-pyridyl
ligands (L), functionalized with
an iodotetrafluorobenzene
moieties, and twelve Pd(II) ions
(M) to undergo, in solutions, a
quantitative self-assembly process
and to form discrete nanocages
of general formula M12L24. Artificial
self-assembled cages can often
encapsulate guest molecules
and promote unusual reactivity;
the XB donor groups decorating
the inside face of the obtained
nanosized capsule will be used
to control the nature of the guest
molecule(s) encapsulated in the
cage and to direct the reactivity of
In the second part of the thesis,
HB is employed in selective
recognition and effective
separation of industrially
important dicarboxylic acids
from either their mixtures
through solid or solution phase
processes. The size-matching of
the interacting partners plays a
major role in allowing for selective
self-assembly and ensuing
separation process. Specifically,
we have demonstrated that bis-
(trimethylammonium) alkane
diiodides, a well-known class of
porous organic salts, can reversibly
encapsulate dicarboxylic acids
through intermolecular hydrogen
―
bonding between the host I
anions and the guest carboxylic
OH group. The process is highly
effective for separating in pure
form of dicarboxylic acid chain that
―
forms an I ···HOOC–(CH2/CF2)n–
―
COOH···I superanion matches
in length to the chosen dication.
The size-matching controlled
Fig. 1 - Crystal structure of supramolecular size matching complexes: Top:
decamethonium iodide adipic acid complex; Bottom: decamethonium iodide
octafluoro adipic acid complex; Colour codes: grey, carbon; blue, nitrogen; white,
hydrogen; red, oxygen; green, fluorine; magenta, iodine.
the selective solubility variation
of dicarboxylic acids and provides
a new direction for selective
recognition and separation.
The strategy is reminiscent of
the effective exchange of α,ω-
diiodoperfluoroalkanes via the
formation of halogen bonded
― ―
I ···I(CF2)nI···I adducts where
cation/supramolecular anion size
matching plays a key role.
In the third part of the thesis
XB is used to form a great
variety of supramolecular
architectures. Specifically,
naked halide anions have been
used to form several halogen
bonded networks with different
and fascinating topologies.
A library of supramolecular
anionic networks showing
Borromean interpenetration
has been prepared by self-
assembly of crypt-222, several
metal or ammonium halides,
and five bis-homologous
diiodoperfluoroalkane, the cation, liquid crystalline elastomeric
or the anion has a different actuators.
relative effect on the metrics and
bondings of the self-assembled
systems, it can be generalized
that bonding, namely energetic,
features play here a less influential
role than metric features in
determining the topology of
the prepared tetra-component
cocrystals. This conclusion
may hold true for other multi-
component systems and may
function as a general heuristic
principle when pursuing the
preparation of multi-component
Fig. 2 - Left: Schematic view of discrete Borromean ring. Right: Partial view
systems having the same topology
(Mercury 3.8, space-filling) of the three honeycomb nets present in the cocrystal
but different composition. crypt-222 potassium iodide diiodoperfluorooctane. Red, yellow, and blue
colours differentiate the three translationally related nets showing Borromean
- entanglement. Supercations K+Ì1 are omitted for clarity.
In addition to this, the role of I···I
XB is used to assess the proton
localization in the product that
hydrogen iodide forms with crypt-
111, a proton sponge with unique
and useful protonation kinetics.
Finally the synthesis and
characterization of new trimeric
complexes obtained upon
XB driven self-assembly of
1,4-diiodotetrafluorobenzene
or α,ω-diiodoperfluoroalkanes,
acting as XB-donors, with an
alkoxystilbazole derivative
functionalized with a methacrylate
terminal group , acting as Fig. 3 - Crystal packing of complexes alkoxystilbazole diiodotetrafluorobenzene
XB-acceptor has examined in (top) and alkoxystilbazole diiodooctafluorobutane (bottom) showing a clear
detail. Despite the fact that segregation between hydrocarbon and perfluorocarbon modules in complex.
Hydrogen atoms have been omitted for the sake of clarity. Color codes: grey,
the starting materials are not carbon; blue, nitrogen; magenta, iodine; red, oxygen; yellow, fluorine.
 346
Recovery of sugars from lignocellulosic
hydrolysates by continuous ion exclusion
chromatography for the production of green
chemicals
Gabriele Lodi - Supervisor: Prof. Laura Annamaria Pellegrini
Tutor: Prof. Davide Moscatelli
Today, the depletion of fossil
resources, the increasing
greenhouse gas emissions and
the resulting climate change,
joined with a growing world
population, make imperative to
find processes able to produce
fuels and chemicals from
alternative, renewable resources.
Lignocellulosic biomasses, e.g.
forestry wastes, wheat straw, corn
stover and sugarcane bagasse,
can be used to obtain monomeric
sugars, which are valuable raw
materials and can be subsequently
converted into fuels or chemicals
through chemical or biochemical
processes.
Biochemical conversion
of biomasses involves
pretreatment and hydrolysis
processes for deconstructing
the recalcitrant lignocellulosic
matrix and for hydrolyzing
the polysaccharides (namely,
cellulose and hemicellulose) to
monosaccharides like glucose,
xylose, arabinose, galactose and
mannose. These monosaccharides
are then fermented or catalytically
converted into the desired
products. Biochemical processes
are often referred to as “sugar-
platform” conversions: the
production of monomeric sugars
from lignocellulosic biomass is
thus the key to a sustainable,
renewable chemical industry.
However, pretreatments of
lignocelluloses release not
only the desired pentose and
hexose sugars, but also various
compounds that can inhibit the
fermentation microorganism
or poison the conversion
catalyst. This can deeply reduce
the product yield and process
productivity, severely limiting
the usefulness of the derived
sugars. In order to enhance the
efficiency of sugars conversion,
it is necessary to remove these
inhibitors from the hydrolysate
before the fermentation.
The recovery of sugars
from hydrolysates and the
removal of by-products can
be successfully accomplished
using chromatography. The
separation is based on a
chromatographic technique
known as Ion Exclusion. In this
technique, strong electrolytes are
separated from nonelectrolytes
and weak electrolytes using a
strong ion-exchange resin as
a stationary phase. The strong
electrolytes are excluded from
the resin due to electrostatic
repulsion with the fixed groups,
while the nonelectrolytes and
weak electrolytes are partitioned
between the mobile phase and the
stagnant liquid inside the particles.
No actual ion exchange takes place
during the separation and hence
no chemical regeneration of the
resin is required.
The Italian company Biochemtex
developed the GREGTM process,
in which monosaccharides are
converted into ethylene and
propylene glycols. The process
is based on traditional catalysts
and a straightforward purification
of the sugars is thus essential
in order to increase the catalyst
lifetime. The process involves an
innovative pretreatment, where
the hydrolysate is neutralized to
pH 6 before the chromatographic
separation making all the
electrolytes in solution completely
dissociated and therefore
excluded from the resin pores. The
electrolytes elute unseparated as
a first group in the chromatogram
followed by the sugars, allowing
a baseline binary separation that
is performed in continuous in an
I-SMB process.
In this work, the recovery of
lignocellulosic sugars obtained
with this innovative pretreatment
has been studied.
Experimental investigations
on model systems, aimed to
understand the mechanisms
involved in the separation, have
been carried out at different
scales. The column dynamics has
been analyzed using equilibrium
and pulse elution data measured
in batch columns. A systematic
PhD Yearbook | 2018
2016
347
Industrial Chemistry and Chemical Engineering
investigation of the adsorption representing the main features
behavior of the main components of this separation. The simulation
found in the hydrolysate has tool can be used for the design
been performed and the effects and optimization of the operating
of the main electrolytes on resin conditions of the I-SMB separation
shrinking and sugars adsorption process.
have been studied. Afterwards,
a quantitative description of the
experimental observations was
provided.
On this basis, a dynamic column
model for IEC has been developed
and used to simulate pulse tests
of real hydrolysates. The exclusion
of ions from the resin pores
has been explained in terms of
Donnan Theory, which describes
the partition of charged species
between two phases separated
by a semipermeable membrane.
The equilibrium relationships for
this model have been derived
and implemented in the material
balances. These were written
taking into account changes in
interparticle and intraparticle
porosity due to resin shrinking.
A model for the I-SMB process was
developed coupling the single-
column dynamic model to the
appropriate material balances at
the nodes. The model was then
used to predict experimental
data of continuous separations,
obtained with the I-SMB unit of
the GREGTM pilot plant. The results
were in good agreement with
the experimental data, correctly
 348
ODOUR EMISSIONS FROM NON-ACTIVE AREA
SOURCES
Federico Lucernoni - Relatore: Prof. Ing. Laura M. T. Capelli
During the PhD, the focus was on
the problem of odour sampling
on non-active area sources. The
final goal was to provide and
answer to the question “Which
is the best way to asses odour
emissions from non-active area
sources?”. It is necessary to clarify
that this is a complex open issue.
In the first part of the research
it was necessary to understand
the sources that were the object
of investigation. A first result was
understanding that the previous
distinction between passive
and active area sources was too
simplistic and that it is necessary to
distinguish the non-active sources
in at least three categories, as
among these categories there are
significant differences that can’t
be neglected if the aim is defining
a method to sample and assess
the odour emissions capable of
producing reliable data. Once this
important first result was achieved,
the focus shifted on the description
and characterization of the three
identified kinds of non-active area
sources: semi-passive sources,
liquid passive sources, solid passive
sources. This led to divide the main
problem in three different sub-
problems, and the initial question
into three distinct questions, each
requiring a different procedure in
order to obtain an answer (see the
scheme in Fig. 1).
Fig. 1 - Scheme of the subdivision of the problem.
Semi-passive sources, are defined
like this as they have a proper
endogenous emission flux, even if
it is typically low; such sources are
typically landfill surfaces. For these
sources, the main results obtained
were the definition of the sampling
hood design and the definition
of the sampling method, both
validated via a detailed on-site
study as well as a thorough CFD
analysis (see Fig. 2).
The purely passive sources,
instead, are characterized by
absence of said endogenous
emission flux and emission is due
to volatilization phenomena. Liquid
surfaces are typically represented
by waste water treatment tanks.
For these sources it was defined
a reliable sampling system, and
another important achievement
was understanding how
important intra-phase transport
PhD Yearbook | 2018
2016
349
different types of non-active area
Industrial Chemistry and Chemical Engineering
sources, as it became evident
how not all are the same and they
can’t be treated all in one single
fashion, also how it is necessary
to differentiate them and how
each kind needs to be treated
distinctly. Then, another important
achievement was obtained,
keeping a more engineering
approach to the problem, that is
Fig. 2 - CFD analysis of the behavior of the sampling device adopted for semi-
the understanding of the crucial passive area sources, exploiting the software Fluent®. It is shown here the methane
role - in the case of passive concentration profile in [kmol/m3] inside the chamber during the 10 min static
deposition time, for one of the scenarios inspected during the study.
sources - played by the transport
phenomena are for a precise phenomena and the characteristic
description of the phenomena chemical-physical parameters of
leading to the emission. For solid the source itself.
sources, typically represented by
contaminated soils, it became
evident that they are characterized
by a high complexity and that in
this case more than in any other,
transport phenomena and the
chemical-physical characteristics
of the source itself are crucial for a
proper approach to the problem.
Another important result was
understanding that the methods
usually adopted, normally
assimilating liquid sources to the
solid ones, neglecting the transport
phenomena and the chemical-
physical aspects, are not proper.
This difference from before and
after this research project is a big
change and it is fundamental.
In conclusion, one of the most
important results of the work,
is the distinction between the
 350
Functionalized hydrogels for drug and cell
delivery in central nervous system
Emanuele Mauri - Supervisor: Prof. Filippo Rossi
Central nervous system (CNS)
injuries represent one of the
most critical scenarios for human
health, society and economy.
Nowadays, the lesions to brain
or spinal cord are still looking for
efficient and sustainable solutions.
In particular, in spinal cord injury,
the neurodegeneration is related
to the influx of immune cells
and the scar formation, which
generates permanent interruption
of information transmission to
and from the CNS. Researchers
are moving towards therapeutic
approaches based on the use
of biomaterials to reduce the
post-traumatic degeneration.
These tools involve the use of
anti-inflammatory agents and
cells to counteract the adverse
effects and to rebuild cellular
interconnections. Recent literature
is focused on in situ treatments
applying three-dimensional
hydrogel scaffolds, due to their
properties: the ability to retain
water mimicking living tissues,
their high biocompatibility, the
possibility to allow controlled
release rates and control the cell
fate.
In this work, functionalized
hydrogel systems, composed by
synthetic and natural polymers,
are synthetized as hospitable
cell delivery and as carriers for
controlled drug release. The
functionalization strategies are
Fig. 1 - In vitro hMSCs viability and morphology. A: cell viability profiles; B-C:
hMSCs in unmodified hydrogel; D-E: hMSCs in hydrogel with RGD and ECM.
post-polymerization modifications
grafting chemical groups that
are not present in the polymer
structures.
Hydrogels are obtained as bulk
systems and nanostructures
(nanogels). The first ones are
developed as appealing cell niches
linking RGD tripeptide to the
scaffold through click chemistry
and introducing the extracellular
matrix (ECM) deposition: the
in vitro and in vivo studies with
human mesenchymal stem cells
(hMSCs) have showed a significant
improvement in cell viability and
density (Fig. 1), promoting the
activation of anti-inflammatory
and curative actions in the target
site.
About the drug delivery guidelines,
the use of pH-sensitive hydrazone
and ester linkers between
drugs and polymer gives the
opportunity to synthetize tunable
tools with different release
kinetics, to promote a sustained
and prolonged drug release. At
different pH, with hydrazone and
ester functionalizations, drug
release has occurred as double
Fickian diffusion regime (Fig. 2):
the transition and duration of
the two regimes depends on the
nature of the functionalization
and probably correspond to ester/
hydrazone lability at the interface
hydrogel/external environment
and ester/hydrazone bonds
present in the inner core.
Moreover, the versatility of
functionalization techniques
allows the design of a polymeric
scaffold as injectable tool for
magnetic resonance imaging
(MRI). In this case, the hydrogel
with paramagnetic properties
is obtained through azide-
alkyne cycloaddition between
polyethylene glycol and TEMPO.
PhD Yearbook | 2018
2016
351
In vivo results have showed the
Industrial Chemistry and Chemical Engineering
traceability of MRI hydrogel signal
at least 1 month, without any
adverse reactions.
On the other hand, the design
of nanogels is aimed to monitor
and interfere in the activation of
inner-cell inflammatory response:
in particular, this work has
investigated the modulation of
nanogel cellular internalization
and their use as drug carrier within
the cytosol. The development
of different polymer-based
coating layers has led to the
modulation of cell uptake, laying
the milestone for the advance
of nanogels capable of selective Fig. 2 - A-C: In vitro release profiles and slope representative of Fickian diffusion
cell interactions: a promising coefficient from unmodified hydrogel (•) and ester-hydrogel at pH 7.4 (♦, blue) and
at pH 8.5 (♦, red). B-D: In vitro release profiles and slope representative of Fickian
approach to introduce targeting diffusion coefficient from unmodified hydrogel (•) and hydrazone-hydrogel at pH 7.4
curative therapy, avoiding (▲, blue) and at pH 8.5(▲, red).
side-effects. The combination
of orthogonal chemistry and loaded. As alternative, the systems, with the idea to improve
physical properties has offered modifications in the experimental their healing effects in the central
the opportunity to synthetize conditions of nanogel synthesis nervous system disorders.
nanogels for drug delivery allows tuning the amount of The combinatorial delivery of
releasing their therapeutic cargo drug loaded and released from drugs and cells through tailored
within the cellular environment. It the nanosystems, showing their biomaterials is a promising and
is proposed an innovative nanogel potential use as drug delivery innovative approach that may
where the drug is grafted to the without constraints related to represent a breakthrough in the
nanostructure through a disulfide the presence of drug functional treatments, with the potential for
bond: after the cellular uptake, groups, otherwise necessary rapid translation to patients and
the drug is released without any to form a chemical bond with a contribute to the advancement
external stimuli or inflammatory polymeric network. in the design and construction of
conditions, but exploiting the In summary, this work, through biomedical devices.
cellular components cysteine and the interdisciplinary approach
glutathione able to disrupt the involving chemistry, engineering
disulfide bond and make available and biology, aims at developing
the bioactive element previously smart drug and cell delivery
 352
Development of Kinetic Mechanisms for the
Combustion of Renewable Fuels
Matteo Pelucchi - Supervisor: Prof. Tiziano Faravelli
Pursuing a sustainable energy
scenario for transportation
requires the blending of fuels
from renewable sources
(alcohols, methylesters etc.) into
hydrocarbon fuels from fossil
sources (gasoline, diesels, jet
fuels etc.). In fact, while effective
alternatives to combustion
exist for electricity production
(nuclear, hydroelectricity, solar,
wind etc.), the high energy
density required for road, sea
and air transport endorses
biofuels as the only viable and
realistic option. Moreover, from
an environmental perspective,
undeniable and dramatic climate
change phenomena impose to
act with long term sustainable
solutions for reducing greenhouse
gases, NO and soot emissions.
x theoretical studies and with experimental measurements from the literature.
If on one side the production of
biofuels from biomasses satisfies
the requirement of a net zero-
CO2 balance, new fuels and
new engines technology have
been investigated to improve
fuel economy and reduce
pollutant emissions. The correct
characterization of a fuel or
fuel mixture reactivity and the
evaluation of its compatibility with
existing engine infrastructures in
terms of pollution and efficiency
is, for a major part, a chemical
kinetics problem.
Combustion kinetic modelling,
which is the topic of this thesis, has
Fig. 1 - Development and validation procedure of the chemical kinetic mechanisms
for pyrolysis and combustion of hydrocarbon fuels.
been driving fuels and engines
development for the last ~25
years. Figure 1 schematically
shows the development and
validation procedure of a kinetic
mechanism according to which
this thesis has been developed.
While most of the focus in the
first decade has been devoted
to the understanding of alkanes
chemistry, the need of better
representing commercial fuels
by means of surrogate mixtures,
extended the interest to aromatics
and lastly to oxygenated fuels
from renewable sources. This
thesis extended the knowledge of
the chemistry involved in alkanes
(e.g. n-heptane), aromatics (e.g.
toluene) and oxygenated (e.g.
alcohols, aldehydes, ketones etc.)
fuels combustion, by means of an
effective coupling of theory and
experiments, within the CRECK
group approach to combustion
kinetics developed in the last
30 years of research activity
at Politecnico di Milano. The
left panel of Figure 3 shows a
comparison between experimental
measurements of the reaction
OH+toluene and results from
theoretical calculation presented
in this thesis. Also the impact of
the updated rate constants on
ignition delay times is reported in
the right panel.
PhD Yearbook | 2018
2016
353
The model obtained coupling Perspectives concerning the
Industrial Chemistry and Chemical Engineering
the different revised portions, possibility and potentials of
accurately reproduces recent effectively and extensively
experimental measurements of exploiting theoretical kinetics, and
surrogate mixtures representative the necessity of fully automatized
of real gasoline fuels. Ignition delay the iterative process of kinetic
time measurements are compared mechanisms development are also
with model predictions in Figure 3 discussed, providing directions for
for mixtures of butanol/heptane future research efforts.
(left panel) and n-heptane/iso-
octane/toluene/butanol (right
panel).
Beside the definition of key
model parameters and standard
kinetic mechanisms validation
procedures, other important
challenges of modern kinetic
modelling have been a topic
of research. The necessity of
automatically assessing the validity
of increasing complexity kinetic Fig. 2 - Left panel: comparison of the total rate constant of OH+toluene with other
mechanisms has been tackled, Right panel: Impact of the update rate constant on toluene/air ignition delay
providing an innovative and time predictions (solid line). Dashed line: POLIMI mechanism with different rate
effective method, of application constants for OH+toluene
also to mechanism reduction and
optimization, and to experimental
design.
Inputs for future refinement
of the presented models have
been provided throughout
the discussion, highlighting
for example the need of a
hierarchical revision starting
from the core C -C portion
0 4
of the POLIMI mechanism or a
better assessment of other key Fig. 3 - Left panel: Ignition delay times for n-butanol/heptane blends at p=20 bar
and φ=0.4. Right panel: Ignition delays for TRF on blending with 20% n-butanol,
channels, whose parameters still stoichiometric TRF and n-butanol mixtures at p = 20 bar in Rapid Compression
carry high degrees of uncertainty. Machine. Symbols: experimental data, lines: POLIMI mechanism.
 354
Halogen Bonding as a new supramolecular
tool to control protein and peptide
self-assembly
Andrea Pizzi - Supervisor: Prof. Pierangelo Metrangolo
Although the abundance of
halogens in the environment is not
remotely comparable to that of
other elements, a remarkable
number (more than 5000) of
naturally occurring halogenated
compounds has been discovered
to date. Halogenation is a minimal
structural modification; however, it
can induce deep changes in the
molecular properties because of
the different chemical features
characterizing halogen atoms.
Among these features, their
tendency to act as electrophilic
species is by far the most studied.
The strong, specific and directional
non-covalent interaction resulting
from this electrophilic behavior is
halogen bonding (XB). Although
halogens are among the most
electronegative elements, they can
act as electrophilic species
because of their anisotropic
distribution of the electron density
around their nucleus. When
halogen atoms form covalent
bonds, the new chemical
environment affects the halogen
electron density, which is no more
equally distributed in all directions.
The electrostatic potential of the
halogen atom becomes
anisotropic, showing distinct
regions around the nucleus with
different chemical behavior. The
external region located along the
extension of the covalent bond
results in lower electron density.
Differently, the region
perpendicular to the covalent
bond shows increased electron
density. The resulting shape of the
halogen atom is an ellipsoid
having the shorter radius (named
polar flattening) along the covalent
bond direction. According to this
electrostatic configuration,
nucleophiles form linear
interactions (160°-180°) respect to
the C−X bond, while electrophiles
give rise to lateral contacts
(90°-120°) against the bond axis.
The strength of halogen bond can
be tuned upon changing the
nature of the halogen atom and its
covalently bound residues.
Increasing halogen atom
polarizability (decreasing
electronegativity) and raising
electron-withdrawing propensity
of its covalently bound moiety
result in a stronger XB.
Consequently, halogen bond
strength increases following the
trend I > Br > Cl >> F. In general,
strong – hence short – interactions
are more directional than long
ones, thus even XB directionality
follows the same trend depending
on halogen atom polarizability,
with iodine that is the more prone
to be involved in halogen bonding.
Because of its peculiar properties
of strength and directionality,
halogen bonding has become one
of the most exploited interactions
in supramolecular chemistry and
crystal engineering. However,
extensive studies focusing on XB
relevance in biomolecular systems
started quite recently. Since it is
known that halogen atoms can
improve biochemical properties
like membrane permeability and
half-life, their introduction in
potential drug candidates is a
common procedure of
pharmaceutical optimization.
Although the importance of
halogens in biomolecular systems
is longer established, a molecular
interpretation correlating the
interaction pattern of halogen
atoms with the chemical
properties of biomolecules is
unusual. A recent survey on the
Protein Data Bank (PDB) nicely
rationalizes XB in biological
systems, showing the more
frequently occurring halogen bond
acceptors. Considering the 567
halogen bonds resulting from this
survey, 430 involve halogen atoms
and protein residues. All of the
twenty amino acids form C−X•••Y
halogen bonds, with leucine as
favorite residue. XBs involving
amino acid main chains are the
majority (64.6%) respect to the
ones formed with side chains
(35.4%). Considering C−X•••π
halogen bonds involving aromatic
amino acids, nearly half (46.4%)
are C−Cl•••Tyr contacts while
22.3% are C−Cl•••Phe interactions.
This thesis highlights the role of XB
in the self-assembly of
biomolecules containing
halogenated phenylalanine (Phe)
residues. The first part of the
thesis describes halogenation as a
tool to improve the self-assembly
properties of a phenylalanine-
based organic gelator: N-Fmoc-
Phe. The hydrogel rigidity of mono
halogenated derivatives of
N-Fmoc-Phe (Fmoc-4-X-Phe, where
X = F, Cl, Br, I) has been related for
the first time with halogen atom
polarizability, i.e. the propensity to
act as halogen bond donor. In
particular, iodinated and
brominated derivatives (containing
the most polarizable substituents)
were found to form the strongest
gels. This trend is opposite to the
one reported in previous studies,
where hydrogel strength increases
with halogen electronegativity.
Kinetic studies confirm that
iodinated and brominated
compounds have the highest
fibrillation propensity, indicating a
greater efficiency to assemble into
high ordered and compact
materials. Gel strength and
self-assembly efficiency are due to
the contribution of halogen
bonding in driving the aggregation
process, as demonstrated by the
crystal structures of Fmoc-4-I-Phe
and Fmoc-4-Br-Phe, showing
halogen bond between the
halogen atom and the electron
density of Fmoc aromatic moiety.
The second part of the thesis
shows the impact of halogenation
on the self-assembly behavior of a
more complex system. The
peptide KLVFF, core sequence of
the amyloid beta (Aβ) protein, was
modified at the para position of
the phenylalanine residues to
obtain seven different
halogenated derivatives. It was
found that some of the
halogenated peptides showed an
increased rate of fibrillation
compared to the wild-type
sequence. In general, iodinated
derivatives self-assembled into
more stable nanostructures, as
confirmed by rheological
experiments showing the
mechanical properties of the
peptide hydrogels directly
connected with halogen
substituent polarizability. Similar
results were obtained with
another amyloidogenic sequence
(DFNKF) although there was a lack
of structural evidence about the
role of halogen atoms in the
self-assembly of this segment. In
this thesis, the crystal structure of
some of the halogenated KLVFF
peptides is the first, direct
demonstration that halogen
interaction pattern nicely
correlates with the aggregation
properties of the peptides. The
crystal structure of the di iodinated
peptide KLVF(I)F(I), showing the
peculiar cross-β spine further
stabilized by halogen bonding
between iodine and peptide
carbonyl oxygens, is consistent
with the strong self-assembly
efficiency of iodinated derivatives.
Brominated and chlorinated
derivatives, forming less compact
aggregates, do not show the
contribution of halogen bonding in
their respective crystal structures.
In addition to the dependence on
halogen atom polarizability, the
self-assembly of the halogenated
KLVFF peptides resulted to be
strongly affected by the number
and position of halogen atoms in
the amino acidic sequence. A
single point mutation like
halogenation was able to induce a
PhD Yearbook | 2018
2016
deep change in the self-assembly
of the peptide. Indeed, TEM
images of each halogenated
derivative showed an impressive
variety of nanostructures. These
results enriches the structural
landscape for peptide self-
assembly, giving access to a new 355
supramolecular tool to control the
Industrial Chemistry and Chemical Engineering
morphology of peptide-based
materials. Considering the
biological context, these results
may also shed light on the impact
of halogenation in vivo, where
oxidative stress mechanisms lead
spontaneously to halogenated
biomolecules. Finally, halogens
were used as “heavy atoms” to aid
the phase determination in X-ray
diffraction experiments. The KLVFF
peptide was modified by placing
iodine and bromine substituents
on the para position of the
terminal phenylalanine ring.
KLVFF(I) and KLVFF(Br) were
successfully crystallized and their
structure was determined with
excellent resolution, by using a
conventional X-ray source. Since
these crystal structures do not
show the halogen substituents
involved in any kind of specific
interaction, they can be considered
a reliable model of the wild-type
sequence, which structure has
never been solved. These crystal
structures confirm the overall
features of the amyloid cross-β
spine, showing in detail the key
non-covalent interactions driving
the self-assembly of the peptide.
For this reason, the crystal
structures of these KLVFF
derivatives may contribute to
enlighten the amyloidogenic
behavior of this deeply studied
core sequence.
 356

Development of hierarchical methodology for limits their application for used to determine the overall heat correlation is implemented in a

PhD Yearbook | 2018

routinely reaction design of
the analysis of novel catalytic reactors: an complex reactor geometries (e.g.
packed bed, foams), which is
application to micro packed bed reactors still based on classical chemical
reaction engineering models.
This type of modeling adopts
Stefano Rebughini - Supervisor: Prof. Matteo Maestri
lumped parameters, usually
experimentally derived, to
The first-principles reactor mainly limited by numerical computational cost. Thus a represent the main phenomena
transfer coefficient between the
micro-packing and the wall of the
single channel of the honeycomb
matrix. Figure 2 illustrates the
comparison between the overall
heat transfer coefficient estimated
with CFD simulations and the
literature correlations.
2016
steady-state, pseudo-continuous
2D heterogeneous reactor
model to analyze the selective
oxidation of o-xylene to phthalic
anhydride in this novel reactor
configuration. This study shows
that the micro packed bed reactor
is characterized by higher heat 357
transfer properties than a packed

Industrial Chemistry and Chemical Engineering

engineering is becoming a very
promising tool for the analysis
of catalytic reactors. It relies on
a fundamental description of
all the phenomena occurring in
the reactor, where each scale
is represented by means of its
own governing equations. In
heterogeneous catalysis, this
approach allows for the accurate
description of the interplay
between chemistry and transport
processes, which is of primary
importance in the understanding
of the macroscopic observed
functionality of the catalyst.
Despite its attractive potential, the
first-principles reactor engineering
is still hampered by numerical
issues and computational
costs, even for simple reactor
geometries. Therefore, it requires
the development of specific tools
and methodologies to enable its
application to advanced reactor
design.
The main aims of this Ph.D. are to
develop a numerical framework
for multi-scale simulations of
catalytic reactors and to propose
a hierarchical methodology for
the design and the analysis of
unconventional and novel reactor
configurations. In fact, despite
its attractive potential, the first
principles multi-scale modeling
of complex reactor geometries
is still impractical. Indeed, it is
issues/limitations related to high
number of variables and the high
computational cost. Therefore,
the first aim of this Ph.D. is the
development a framework,
which can be used for the first
principles multi-scale modeling
of catalytic reactors. In particular,
a new numerical multi-scale
framework able to represent
catalytic reactor with different
geometries has been developed.
This new framework enables
the concomitant description of
transport phenomena in the gas
phase and in the catalyst. It has
been applied to analyze the effect
of the channel cross-section in
the catalytic partial oxidation of
methane in a honeycomb reactor.
In particular, this framework
allows a detail description of the
transport phenomena between
the gas and the catalyst, which has
been used to investigate the effect
of the geometry on the reactor
performance. This analysis shows
that the channel shape strongly
affects the reactor behavior
in terms of temperature and
composition profile. In fact, due
to the different diffusion length of
the two different cross-sections,
a different amount of reactants
reach the catalytic wall, affecting
the reactor performance.
One of the main limitations
of this framework is the high
numerical method to reduce the
computational effort when the
surface chemistry is described
by means of microkinetic
models has been developed.
This methodology enables the
use of complex homogeneous
kinetic schemes coupled with
micro-kinetic heterogeneous
schemes, which are applied
to investigate the adequacy
of lumped parameters in
describing the heterogeneous
and homogeneous chemistries
interaction in the catalytic partial
oxidation of hydrocarbons
fuel in monolith reactors. This
analysis shows that when the
effect of the homogeneous
chemistry is weak the lumped
parameters implemented in
the 1D model are adequate to
describe the reactor behavior.
On the contrary, when the effect
of the homogeneous chemistry
is enhanced by increasing the
pressure, the lumped parameters
are not accurate enough to
correctly describe the interaction
between the surface and the gas
chemistries.
The previous analyses clearly
show the deeper understanding
of the reactor behavior which
can be obtained by using this
multi-scale numerical framework.
However, the high computational
cost of multi-scale simulations
occurring inside the reactor. This study allows the selection bed reactor. In fact, the micro
Therefore, the second goal of this of the most adequate literature packed bed reactor has a quasi-
Ph.D. is to develop a methodology correlations for describing the isothermal behavior.
to enable the application of transport phenomena in micro On one side, this work shows that
the multi-scale framework to packed bed reactors. Then, this micro packed bed reactors are a
the routinely reactor design. valid alternative to the classical
In this regard, the hierarchical multi tubular packed bed reactor
approach has been selected as an for selectivity oxidation. On the
interesting approach to meet this other side, it clearly shows the
goal. In fact, it can be adopted to potentiality of the hierarchical
enable a first principles multi-scale analysis and how it can be
design of the novel reactors with employed for the efficient and
an affordable computational cost. fundamental analysis and design
In essence, as shown in Figure 1, it of novel reactor technologies.
consists in using the first principles
multi-scale model to analyze a
Fig. 1 - Schematic description of the
selected and limited number hierarchical modeling applied to novel
of operating conditions for the reactor configurations
novel reactor geometry. Then,
the results of these simulations
are used to derive lumped
parameters. Consequently,
the implementation of these
parameters in simplified models
enables a fundamental description
of the phenomena in the reactors
with a reasonable computational
cost. The potentialities of this
methodology are presented by
analysis the behavior of micro
packed bed reactors in high
exothermic processes. The
hierarchical modeling is applied
to investigate the capabilities
of micro packed bed reactors
in dealing with high exothermic
processes. In particular, the
hierarchical approach can be also Fig. 2 - Overall heat transfer coefficient as a function of the Reynolds number
 358
Process Systems Engineering for early-
stage process development: the case of
sustainable bio-derived adipic acid
Alessandro Rosengart - Supervisor: prof. Attilio Citterio
The awareness of the negative
impacts of human activity against
environment and public health
has pushed western governments
to support long-term programs
aimed at mitigating pollution and
reducing resource consumption.
In this spirit, both industry and
academia are searching for new
solutions towards a “green”
manufacturing practice, and the
concept of “biorefinery” is taking
place, as a renewable counterpart
of the ill-famed oil industry.
Biorefineries are supposed
to produce entire classes of
chemicals and fuels just as a real
refinery, with the great difference
that the carbon source is no more
fossil, but follows the natural cycle
of CO2, which is captured from
atmosphere and fixed into living
organisms (plants, algae, bacteria).
My Doctoral Thesis deals with
the feasibility evaluation of a
drop-in biorefining application
for the production of sustainable
nd
adipic acid from biomass (2
generation technology), defining
the full-scale process flowsheet,
assessing the environmental and
economic performances, and
identifying the current challenges
that R&D should address before
industrialization. Some of these
challenges have been tackled
in this work applying several
Process Systems Engineering
computational tools, in particular
dealing with problems of
predictive models development,
uncertainty propagation study
and parameter regression from
experimental data.
The current adipic acid production
covers a market of 3.7 million
tons per year (with a 4.1% of
yearly growth) and, in spite of 70
years of technological maturity,
the traditional benzene-based
processes still raises serious safety
and environmental concerns.
For these reasons, both private
and public research institutions
have pursued alternative bio (and
chemical) routes for adipic acid;
still none of these processes has
reached industrialization yet, also
due to the oil-price fall in 2014.
This event evidenced the main
weakness of drop-in biorefineries:
the need to compete in costs with
a well-established and optimized
technology. A novel approach to
process development is therefore
required for the case of bulk bio-
derived chemical with low added
value. In particular, conceptual
design acquires particular
importance from the early stage
of process development, to
produce reliable cost estimates
and projections, and to define a
strategy for R&D.
The first activity carried out
for this Doctoral study was the
collection and systematisation of
the extensive and interdisciplinary
available knowledge on adipic
acid, identifying the current
alternative processing routes
and assessing their actual
sustainability with objective
green metrics. In the specific,
a two step biological-chemical
process was considered worth of
more detailed investigation for
its good yields and sustainability
potential. This process consists
in a first fermentation to produce
an unsaturated intermediate,
muconic acid, starting from either
glucose (from cellulose) or benzoic
acid (from lignin). Muconic acid is
then catalytically hydrogenated to
adipic acid.
A computer aided process
synthesis-and-design methodology
was therefore applied for the case
study of adipic acid from muconic
acid, to evaluate systematically
the highest number of process
alternatives to produce the
best flowsheet concept at the
state of the art. The method is
implemented in a tool belonging
to “ICAS software”, developed at
the collaborating institution DTU
(Denmark Technical University,
Copenhagen). This tool, given a
number of alternative feedstocks,
technologies (intended as unit
operations/process conditions),
and products, allows building a
superstructure, which is translated
into a Mixed Integer (Non) Linear
Programming optimization
®
problem solved in GAMS . The
solution is a processing route that
maximizes the objective function
(economic potential) providing also
the material and energy balances.
The plant feasibility is then
evaluated for different market
scenarios, the process bottlenecks
are identified and more detailed
green metrics are calculated
(e.g. water consumption, energy
consumption, CO2 equivalent).
Thus, the process flowsheet
concept is achieved avoiding the
simplified approaches of order-
of-magnitude estimates or the
“analogy principle”. Also, the
used methodology guarantees
in general higher flexibility than
a detailed process simulator, as
each unit operation is defined
by few user-defined parameters,
that allow giving some cost/
performance estimates even
in presence of preliminary, lab-
scale data. In case of missing
information, some assumptions
can be made, which become
the object of future research, if
proven determinant in the process
economics (SMART objectives
definition).
The problem of how scarce
information at early stage
of research can affect the
trustworthiness of a cost function
was addressed in detail for a
unit operation which relies much
on experimental data: bacterial
cross flow microfiltration. In facts,
using literature data in analogy
with existing plants can be
extremely deceptive, as the design
and the operating conditions
(transmembrane pressure,
crossflow velocity, membrane
regeneration) are calibrated
on the specific properties of a
particular microorganism. A grey-
box model was developed on
the structure of a Darcy additive
resistance equation, to represent
the dynamic behaviour of the
progressive bacterial fouling
of a filtration membrane. The
model, a system of algebraic
and differential equations, is
general and flexible, able to
represent different strains in
virtue of its adaptive parameters.
The parameters “carry” the
uncertainties deriving from the
experimental error or from the
actual representation limits of the
model. Further uncertainties are
introduced when using the model
in a predictive way, i.e. extending
the parameters validity to “similar”
systems, as performed for the
case of the strains for muconic
acid production, whose filterability
properties have not been
measured yet. The uncertainty
propagation was therefore
studied applying Possibility theory
and the Fuzzy Logic of Zadeh.
This uncertainty analysis allows
identifying the most likely range
of filtration performances of an
industrial membrane system,
which, in association with a cost
function, provides an indication of
the error of cost estimates and the
risk in the absence of specific R&D.
The final conversion step for the
production of green adipic acid,
i.e. the catalytic hydrogenation of
muconic acid, lacked as well of the
sufficient data to perform reliable
estimates on the reaction scale
up. In this case, an experimental
campaign was started, in
collaboration with the laboratories
PhD Yearbook | 2018
2016
of Industrial Chemistry of
Università degli Studi di Milano.
The purpose was to identify the
optimal reaction conditions (low
pressure, low temperature, and
catalyst recyclability) and identify
the hydrogenation mechanism,
to develop the first kinetic model 359
for the system. Several models
Industrial Chemistry and Chemical Engineering
with LHHW structure were used
to interpret the experimental
values, considering the species
adsorption-desorption equilibria
of the involved species. A dual-
step hydrogenation mechanism
was demonstrated, with
hydrogen dissociation on the
metal (Pt/C 5%). Muconic acid
(in its trans,trans form) is first
hydrogenated to hexenedioic acid
(present in its two cis and trans
isomers, in equilibrium) which
is then converted to adipic. The
model parameter regression was
performed with the C++ library
BzzMath, characterized by robust
minimization algorithms, to tackle
the computational challenges
related to the use of models with
strong collinearity. The models
were thus re-parametrized and
progressively simplified, obtaining
a good representation of the
experimental data and providing
the first reference values of the
species activation energies.
 360
MULTICOMPONENT AND CASCADE REACTIONS FOR
THE PREPARATION OF HETEROCYCLIC MOLECULES
OF BIOLOGICAL AND CHEMICAL INTEREST
Arianna Rossetti - Supervisor: Prof. Alessandro Sacchetti
In the last century, the strong
demand for new chemical
compounds drove the
development of novel synthetic
strategies for the easy and fast
generation of molecules libraries.
In particular, in the field of
Diversity Oriented Synthesis (DOS),
convergent synthetic methods,
such as the Multicomponent
Reactions (MCRs), have emerged as
powerful tools for this target.
In this context, my PhD research
project has focused on the
synthesis of new chemical
scaffolds through different
multicomponent and cascade
reactions. The approaches
investigated have led to the
production of different classes of
molecules with either biological
or chemical interests. The final
aim was the creation of libraries
of compounds with high degree
of diversity and possible useful
future applications in the fields of
drug discovery, catalysis or new
material sciences.
In details, the synthetic procedures
for the applied reactions are well
established, nevertheless each
type of MCR has been optimised
in its conditions, by testing
different temperatures, solvents
and combination of equivalents
before the creation of the library
of desired compounds.
Concerning the molecules with a
PhD Yearbook | 2018
2016
361
potential pharmaceutical interest, of thiophene[3,2-d]pyrimidine
Industrial Chemistry and Chemical Engineering
the syntheses were assisted and scaffolds and derivatives
integrated by in silico evaluation of as potential PPI targeting
the targets. peptidomimetics (chapter 4a)
Finally, for selected compounds, and in the synthesis of novel
preliminary biological tests tetrahydrothieno[2,3-c]pyridine
towards cytotoxicity and derivatives as new antimicrobial
antimicrobial activities have been agents (chapter 4b).
performed. This is the case of Last, Mannich reactions were
the products from the Gewald used for the creation of highly
reaction, developed as promising functionalized bispidinones, whose
antitumoral and antimicrobial applications vary from material
agents, or the macrocylic scaffolds science, as in the case of our new
from the Ugi reaction, designed to MOFs scaffolds synthesis (chapter
be PPI inhibitors. 5a), to the catalysis world in
enantioselective Henry reactions
After a brief overview of the MCRs (chapter 5b).
world (chapter 1), the thesis is
organized dividing the libraries
of heterocycles synthesized
according to the corresponding
multicomponent or cascade
reaction applied.
In detail, the first reaction
explored (chapter 2) was an
isocyanide-based cascade used
for a new and facile synthesis of
N-substituted-benzimidazoles.
Then the Ugi reaction, a
multicomponent approach for
peptidomimetics synthesis, was
investigated both on 3,4-dihydro-
isoquinolines as reverse turn
mimics (chapter 3a) and on
macrocycles as PPI inhibitors
(chapter 3b).
A series of Gewald MCR
afforded highly functionalized
2-aminothiophenes in the design
 362

Enantioselective synthesis of chiral library, a whole set of variants, was The optimization of enzymatic Two stereogenic centres in 1,2

PhD Yearbook | 2018

considered. Protein engineering
pharmaceutical intermediates by whole cell procedures were employed to
obtain a single mutation in a
microorganisms and engineered isolated enzymes position with a critical role for the
catalytic function of the protein.
The effects of this mutation on
Sara Santangelo - Supervisor: Prof. Maria Elisabetta Brenna
enantioselectivity were deepen in
Tutor: Prof. Carlo Punta relationship with the substrates of
interest.
cascade procedures has involved
coupling more than one enzyme
into an efficient pathway for the
biosynthesis of compounds of
commercial interest. The most
odorous stereoisomers of the
chiral commercial fragrance
Muguesia® are prepared by a
very effective linear biocatalysed
2016
relative position are thus created
under high stereochemical control
by a two-step one-pot enzymatic
procedure.
These procedures represent a
further demonstration of the
synthetic potential of enzyme-
mediated reactions: the high 363
chemo and stereoselectivity that

Industrial Chemistry and Chemical Engineering

The role of biocatalyzed
reactions is becoming more
and more influential in the
toolbox of available reactions
for the synthesis of enantiopure
compounds, also on industrial
scale.
In the scenario of the possible
options to design the synthetic
pathway of chiral molecules,
enzymatic biotransformations
started to be evaluated more
frequently, also considering
the distinctive features of this
methodology. Enzyme catalyzed
reactions are highly chemo, regio
and stereo-selective and therefore
of great interest for fine chemical
synthesis, both for economic and
environmental reasons.
Enzymes catalysed reactions
can be performed in different
conditions: the natural expressing
microorganism can be considered
as source of enzyme and it can
be used directly as biocatalyst.
To improve overexpression
and to reduce side reactions, a
more convenient host can be
identified. The enzymes can work
in a more known and controlled
overexpression system, leading to
better results. Using engineered
procedures, the possibility to
purify the enzymes becomes
available. In this way, they can be
used for biotransformations also
as purified enzymes. All these
options have been considered and
investigated.
The work during these past
three years has explored the
effectiveness of combining
chemical and biochemical
procedures for the synthesis
of chiral building blocks in high
enantiomeric purity. Specifically,
this has involved three different
topics:
• Stereoselective reduction of
C=C and C=O double bonds;
• Desymmetrisation of achiral
compounds by biocatalysed
oxidative biotransformation
• Combination of multiple
enzymes in cascade reactions
A complete investigation of the
steric and electronic effects
of substituents on the course
of ER-mediated reductions of
alkenes has been performed,
considering the substitutions on
aromatic rings and on alkyl chains.
The conversion yields and the
enantioselectivity of the reactions
were then evaluated.
At the same time, the effect of
different electron withdrawing
groups has been investigated: in
order to perform the enzymatically
catalysed reduction of the
molecules, the of C=C double
bond has to be efficiently activated
under the electronic point of view.
The role of the nitrile (CN) and
the nitro (NO ) functions as EWG,
2
either alone or in combination
with other functional moieties
have been considered. The
enantioselective reduction of
alpha-methylenic nitrile derivatives
catalysed by ene-reductases
affords the corresponding (R)-2-
arylpropanenitriles with high
conversion values. The reaction
is investigated either in aqueous
medium (with an organic cosolvent
or by loading the substrate
unto hydrophobic resins), and
in a biphasic ionic liquid / water
system). Compounds bearing
either alkyl chains of increasing
length at the carbon atom in
position β to the nitro group or
different substituents on the
aromatic ring are prepared and
submitted to bioreduction, in
order to define the synthetic
potential of this enantioselective
reaction in the preparation of
chiral fine chemicals. A complete
screening has been performed,
analysing all the factors influencing
the reactions. The synthetic
versatility of the nitrile and nitro
functions were further investigated
with the transformation of
reduced compounds in the
corresponding chiral amides, acids
and amines.
When the wildtype enzyme wasn’t
able to perform the required
biotransformations, also the
possibility of use an enzyme
cascade reaction, in which a enzymes can achieve are key
The study of the desymmetrisation suitable unsaturated ketone is requisites for the optimization
of achiral 1,3-diols by biocatalysed submitted to the sequential action processes, which are now
oxidation has been conducted of two enzymes, an ene-reductase extensively investigated for the
by exploring the scarcely known and an alcohol dehydrogenase, synthesis of valuable compounds.
world of Acetic Acid Bacteria. The which are added together to the
ability of AAB to oxidize several same reaction vessel with the
substrates has long been known cofactor regeneration system.
and is still attracting attention.
The oxidation of alcohols by
AAB is an old and established
microbial method for obtaining
production of vinegar and various
carboxylic acids, sometimes
with remarkable chemo and
enantioselectivity. The enzymatic
oxidation of primary alcohol Fig. 1
is attractive because it can be
carried out under mild conditions
that are also suited for labile
products. Two different enzymes
are involved in two different
steps for the oxidation to obtain
carboxylic acid in physiologic
pathways and conditions: The
first reaction is catalyzed by
alcohol dehydrogenase (ADH)
and the second by aldehyde
dehydrogenase (ALDH). The
procedure, carried out in aqueous
medium under mild conditions of
pH, temperature and pressure,
contributes to enlarge the
portfolio of enzymatic oxidations
available to organic chemists for
the development of sustainable
manufacturing processes.
Fig. 2